-- r -,

Witchcraft, Oracles,
and Magic among
~1 the Azande
•
.~
~1 ~
'I
I,~

E. E. EVANS-PRITCHARD

ABRIDGED WITH AN INTRODUCTION

BY Eva Gillies

.'.
1

,1

CLARENDON PRESS· OXFORD

197 6
 1
CHAPTER VIII
The Poison Oracle' in Daily Life
I out of the paste into the beaks of small domestic fowls which
ORACLES are a more satisfactory means of ascertaining the
future, and hidden things of the present, than are witch-doctors.
Witch-doctors are useful as sleuths for seeking out the many
affairs of a group of homesteads, and their chief value is that
they generally clear the atmosphere of witchcraft. On this
account they are often asked to dance before a big hunt because
this is ajoint undertaking, many persons are involved, and the
interests of a district are at stake. A public attack by witch-
doctors, who act as ritual skirmishers to report on and to
counter the mystical forces in opposition, is appropriate. When
the seance is over people feel that witches have been scared from
their undertaking.
But as diviners witch-doctors are not regarded as furnishing
more than preliminary evidence, and in all matters of moment
a man takes a witch-doctor's statement and places it before one
of the greater oracles for corroboration. This is, moreover,
necessary if a man wishes to take any public action. He cannot
try to exact vengeance for homicide on the evidence of a witch-
doctor alone. A witch-doctor would never be consulted on such .
a matter. A man would be very ill-advised even to present a
fowl's wing to a witch accused solely by witch-doctors. The
accused might mock the bearer of the wing and would not lose
esteem for doing so. Hence Azande say that witch-doctors, like
the rubbing-board oracle, are useful because they can answer
quickly many questions and sort out suspects in a preliminary
manner before men approach the poison oracle, but that they
are not dependable.
II
The method of revealing what is hidden by administering
poison to fowls has a wide extension in Africa; but just as the
Azande are the most north-easterly people who have the notion
Oracles 121
ofwitchcraft as a material substance in the belly, so also is their
culture the north-easterly limit of the distribution of this type
oforacle. They are the only people in the Anglo-Egyptian Sud-
dan who employ it.
The poison used is a red powder manufactured from a forest
creeper and mixed with water to a paste. The liquid is squeezed
are compelled to swallow it. Generally violent spasms follow.
The doses sometimes prove fatal, but as often the fowls recover.
Sometimes they are even unaffected by the poison. From the
behaviour offowls under this ordeal, especially by their death
or survival, Azande receive answers to the questions they place
before the oracle.
The botanical nature of the poison has not been determined,
but its chemical nature has been roughly analysed. Some of the
oracle poison which I brought back to England was examined
by Professor R. Robinson who informs me that:
The quantity of benge was insufficient to enable me to establish with
certainly the nature of the active principle. All that can be said about
it is that the toxic substance is alkaloidal in character and appears
to be related chemically to strychnine. It is almost certainly not homo-
geneous, and this accounts for the difficulty ofisolation in a pure con-
dition. Thus, all I can say is that it is strychnine-like in many of its
reactions, and that probably two or more bases are present.
III
The poison oracle,)~~ngi; iLh}'Jax-th.e...m.ost-importan.L-aLthe
Zande oracles. Zande rely completely on its decisions, which
have the force of law when obtained on the orders of a prince.
A visitor to Zandeland hears as much of the poison oracle as
he hears of witchcraft, for whenever a question arises about the
facts of a case or about a man's well-being they at once seek
:
to know the opinion ofthe poisonoracle on the matter. In many
situations where we seek to base a verdict upon evidence or try
I'
jl'
...~
to regulate our conduct by weighing of probabilities the Zande
consults, without hesitation, the poison oracle and follows its i
l
I'
directions with implicit trust. '~ ,
eNo important venture is undertaken without authorization :~
:I~
of the poison oracl~ In important collective undertakings, in
all crises of life, in all serious legal disputes, in all matters .• ~~
jj
.)~1
,1
. ~~j'
.,
 I
122 Oracles
strongly affecting individual welfare, in short, on all occasions
regarded by Azande as dangerous or socially important, the
activity is preceded by consultation of the poison oracle.
I do not wish to catalogue all situations in which the oracle
may be consulted since this would mean a list ofsocial situations
in every sphere ofZande life, and when each sphere is described
the part played by oracles is more fitly recorded than in the
present place. Notwithstanding, it is desirable to list some of
the occasions on which the oracle must be consulted in order
to give the reader a clear idea of its significance to Azande.
When I say that the poison oracle, or some other oracle, must
be consulted on the occasions listed below, I mean that if a
Zande were not to consult it he would be acting contrary to
custom and might suffer in social prestige. He might even incur
legal penalties. The following situations are typical occasions
of consultation:
To discover why a wife has not conceived.
During pregnancy of wife, about place ofdelivery, about her safety
in childbirth, and about the safety of her child.
Before circumcision of son.
Before marriage of daughter.
Before sending son to act as page at court.
In sickness ofany member offamily. Will he die? Who is the witch
responsible? etc.
To discover the agent responsible for any misfortune.
At death of kinsman in the old days. Who killed him? Who will ,
execute the witch? etc.
Before exacting vengeance by magic. Who will keep the taboos?
Who will make the magic? etc.
In cases of sorcery.
In cases of adultery.
Betore gathering oracle poison.
Before making blood-brotherhood.
Before long journeys.
A man before marrying a wife.
Before presenting a prince with beer.
Before large-scale hunting.
A commoner in choosing a new homestead site.
Before accepting, or allowing a dependant to accept, European
employment.
Before becoming a witch-doctor.
Oracles 12 3
Before joining a closed association.
A man before he and his adult sons go to war.
In cases of disloyalty to a prince.
A prince before making war.
To determine disposition of warriors, place and time of attack, and
all other matters pertaining to warfare.
A prince before appointing governors, deputies, or any other
officials.
A prince before moving his court.
A prince to discover whether a communal ceremony will terminate
drought.
A prince to determine the actions of the British District Commis-
sioner.
A prince before accepting presents and tribute.
IV
It is not only about what we would consider the more important
social activities that Azande consult their oracles, but also about
their smaller everyday affairs. If time and opportunity per-
mitted many Azande would wish to consult one or other of the
oracles about every step in their lives. This is clearly impossible,
but old men who know how to use the rubbing-board oracle
usually carry one about with them so that if any doubt arises
they can quiet it by immediate consultation.
A typical occasion on which a man consults his rubbing-
board oracle is when he is on a visit to a friend's homestead.
When his visit is concluded he asks the oracle whether he had
better leave openly during the daytime or depart secretly at
night so that any witch who may wish to dispatch his witchcraft
after him, to cause him some misfortune on the journey, may
be ignorant that he has left. If the oracle advises him to depart
at night he tells his host and leaves before dawn. Other members
of the homestead understand what has happened and are not
angry that he has not bid them farewell. Or the rubbing-board
oracle may tell a man that he can depart in the daytime but
must be careful about witchcraft on the way. In this case he .1'
,I
strolls away from his host's homestead as though he were taking .\
a short walk and throws a spear-shaft aimlessly in front of him
so that people who observe him on the path' think that he is :'i
playing and will shortly return from his stroll, since people J
,rl!
departing on a journey do not meander at the start. When he 1~
124 Oracles
is well out ofsight he quickens his steps and hastens on his way.
Sometimes he does not even inform his host of his departure,
but the host understands the reason for his silence.
I found that when a Zande acted towards me in a manner
that we would call rude and untrustworthy his actions were
often to be accounted for by obedience to his oracles. Usually
I have found Azande courteous and reliable according to
English standards, but sometimes their behaviour is unintelli-
gible till their mystical notions are taken into account. Often
Azande are tortuous in their dealings with one another, but
they do not consider a man blameworthy for being secretive
or acting contrary to his declared intentions. On the contrary,
they praise his prudence for taking account ofwitchcraft at each
step and for regulating his conduct after the direction of his
oracles. Hence it is not necessary for one Zande to explain to
another his waywardness, for everybody understands the
motives of his conduct.
Not all Azande are equally prone to consult oracles. I have
frequently observed that some men are more keenly aware of
danger from witchcraft than others and rely far more than
others upon magic and oracles to counteract its influence. Thus
while some men like to consult oracles and to blow magic
whistles or perform some other magic rite before embarking
upon even small adventures, other men only consult oracles
about important legal issues and at real crises, such as
marriage, serious sickness, and death. When they are socially
compelled to consult oracles they do so, but not otherwise. '
In legal procedure everyone must make use of the poison
oracle.
To understand Zande legal procedure one must know exactly
how the poison oracle is operated, because in the old days it
was in itself the greater part of what we know as rules of evi-
dence,judge,jury, and witnesses. In the past the two main types
of cases were witchcraft and adultery. Witchcraft cases were
settled entirely through the oracles since there was no possibility
of discovering mystical action except through the mystical
power ofthe poison oracle. All a prince had to do was to confirm
the names ofwitches discovered by the kinsmen of dead persons
by placing their names before his own oracle. The compensa-
tion which a witch had to pay for his crime was" fixed by custom.
Oracles 12 5
All death to Azande is murder and was the starting-point of
the most important legal process in Zande culture. Azande
therefore find it difficult to see how Europeans can refuse to
take cognizance ofwhat is so manifest and so shocking to them.
In a case of adultery there might be circumstantial evidence,
but in fact simple cases of this kind were rare. The chance dis-
covery of lovers during a few minutes' congress in the bush or
during the absence ofa husband from his homestead was small.
The only certain evidence upon which a suspicious husband
could act was that provided by the poison oracle, for even if
a wife repented of her infidelity and told her husband the name
ofher lover he might deny the accusation. The husband might,
it is true, urge before the prince some other grounds for suspi-
cion, but he would base his charge of adultery mainly upon
the evidence of the oracle, and no further proof than this was
required. The accused man would defend himselfless by urging
absence of circumstantial evidence than by offering to give a
ngbu or test. He was asked to choose a man of substance among
the regular attendants at Court and to give him the test, telling
him to place the question of adultery before his oracle. This
man acted on behalf of his prince and the declaration of his
oracle settled the case. To Zande eyes this is the perfect pro-
cedure in adultery cases and they do not approve of Eu'ropean
methods, for in their opinion the only sure evidence of guilt
or innocence is not allowed.
Accusing husbands and men accused share this opinion, the
husbands because they often cannot produce evidence accept-
able to government courts of adulteries for which they possess
conclusive proofin the declaration ofthe poison oracle; accused
persons, because they are condemned on the declaration of a
woman without appeal to the one really reliable authority, the
poison oracle.
Special care is taken to protect a prince's oracle poison from
witchcraft and pollution because a prince's oracles reveal mat-
ters of tribal importance, judge criminal and civil cases, and I~~l
determine whether vengeance has been exacted for death. A ,;:
prince has two or three official operators who supervise his
poison oracle. These men must be thoroughly reliable since the 'I
~,\
fate of their master and the purity of law are in their hands.
If they break a taboo the whole legal system may become ~',;
;;;:1
·1;
'I'\~
.....oii
126 Oracles
corrupted and the innocent be judged guilty and the guilty be
judged innocent. An official consulter of a prince's oracles must
also be a man ofimpeccable honesty since he is given sole charge
ofmany legal cases and tests of vengeance. He can ruin subjects
of his master by fabricating oracular statements. Finally, the
consulter ofa prince's oracle must know how t6 maintain silence
about his master's affairs. There is no offence more serious in
the eyes of a Zande prince than 'revealing the speech of the
king's poison oracle'.
We who do not believe in the poison oracle think that the
courts we have established are just because they recognize only
evidence which we regard as such, and we flatter ourselves that
they are native courts ofjustice because we allow natives to pre-
~
side over them. But Azande think that they do not admit the
only evidence which is really relevant to the cases which come
before them, and the princes who have to administer justice
do so with mechanical application of imported European rules
of procedure, and without conviction, since the rules are not
according to custom.
v far removed from homesteads. Any place in the bush screened
I never found great difficulty in observing oracle consultations.
I found that in such matters the best way of gaining confidence
was to enact the same procedure as Azande and to take oracular
verdicts as seriously as they take them. I always kept a supply
of poison for the use of my household and neighbours and we
re(ulated our affairs in accordance with the oracles' decisions:
-I may remark that I found this as satisfactory a way of running
my home and affairs as any other I know of. Among Azande
it is the only satisfactory way of life because it is the only way
oflife they understand, and it furnishes the only arguments by
which they are wholly convinced and silenced. Friends and
neighbours would from time to time ask me to let them bring
fowls to COI).sult my oracles about their troubles. I was always
pleased at this sign of their trust. Also, I had opportunity on
a number ofoccasions to observe Qther people's oracles at work.
In the course of many.months I made. repeated observations
of oracular consultations and had ample opportunity to
acquaint myself with details of technique and interpretation.
An investigation into the use of the poison oracle, like an investi-
Oracles 127
gation into beliefs about witchcraft, does not require special in-
formants. I could rely upon direct observation and could elicit
commentary from any adult Zande when a point was not
wholly clear to me. I can say the same about the rubbing-board
oracle and, to a lesser degree, of the termites oracle.
For information on the following points, however, I had to
rely mainly, or entirely, on verbal information: the process of
collecting oracle poison; the administration of poison to human
beings; and the use of the poison oracle in judicial procedure
at the king's court. Poison is not administered to human beings
at the present time. The poison oracle has no longer a primary
role in court procedure, though it is still to some extent
employed. It had been my ambition to observeor:aclepoison
being gathered and I made an expedition into the Belgian
Congo with this end in view but was defeated by combined
dysentery and malaria, and was carried home again in extreme
weakness.
VI
The usual place for a consultation is on the edge of cultivations
by high grasses and brushwood is suitable. Or they may choose
the corner of a clearing at the edge of the bush where crops
will later be sown, since this is not so damp as in the bush itself.
The object in going so far is to ensure secrecy, to avoid pollution
by people who have not observed the taboos, and to escape
witchcraft which is less likely to corrupt the oracle in the bush
than in a homestead.
One does not consult the poison oracle during the heat of
the day since strong sunlight is bad both for the poison and
for the chickens. If the oracle is consulted late in the morning
the basket of chickens is placed in the shadeof a nearby shrub
or covered with grass. When the poison has been for some time
in strong sunlight it becomes very potent and they say then that,
'If a man gives one dose to a small fowl he has given it quite ;:1,
"j';
enough.' The normal time for consultations is from about eight ,t-
1J
to nine o'clock in the morning, because by this time the dew 'I
I
"
has evaporated and it is possible to sit down in the bush without :1
great discomfort. Very occasionally elders who frequently con- !~
sult the oracles and conduct long seances hold them at night.
"",
 128 Oracles
The consultation may then take place in the centre of the
homestead after the womenfolk have retired to bed. Consulta-
tions may take place on any day except the day after a new
moon.
Oracle poison is useless unless a man possesses fowls upon
which to test it, for the oracle speaks through fowls. In every
lande household there is a fowl-house, and fowls are kept
mainly with the object of subjecting them to oracular tests. As
a rule they are only killed for food (and then only cocks or old
hens) when an important visitor comes to the homestead, per-
haps a prince's son or perhaps a father-in-law. Eggs are not
eaten but are left to hens to hatch out. Clay receptacles may
be fashioned or baskets placed in one of the huts to encourage
hens to nest in them, but often they lay their eggs in the bush
and if they are fortunate will one day strut back to the home-
stead accompanied by their broods. Generally a lande, unless
he is a wealthy man, will not possess more than half a dozen
grown fowls at the most, and many people possess none at all
or perhaps a single hen which someone has given to them.
Small chickens, only two or three days old, may be used for
the poison oracle, but Azande prefer them older. However, one
sees fowls of all sizes at oracle consultations, from tiny chickens
to half-grown cockerels and pullets. When it is possible to tell
the sex offowls Azande use only cockerels, unless they have none
and a consultation is necessary at once. The hens are spared
for breeding purposes. Generally a man tells one of his younger
sons to catch the fowls the night before a seance. Otherwise they
catch them when the door of the fowl-house is opened shortly
after sunrise, but it is better to catch them and put them in
a basket at night when they are roosting. For if the fowls elude
capture in the morning and run away into nearby gardens it
is much trouble to catch them. Two or three boys have to run
them down, all the womenfolk know what is going on, the
neighbours hear the noise, and a witch among them may follow
the owner of the fowls to prevent the oracle from giving him
I the information he desires. When chickens are used this diffi-
culty does not arise because they skep in one of the huts, where
they are immune from attacks by wild cats, and they are easily
caught on the morning of a seance.
Old men say that fully grown birds ought not to be used in
Oracles 12 9
oracle consultations because they are too susceptible to the
poison and have a habit ofdying straight away before the poison
has had time to consider the matter placed before it or even
to hear a full statement of the problem. On the other hand,
a chicken remains for a long time under the influence of the
poison before it recovers or expires, so that the oracle has time
to hear all the relevant details concerning the problem placed
. before it and to give a well-considered judgement.
VII
Any male may take part in the proceedings. However, the
oracle is costly, and the questions put to it concern adult
occupations. Therefore boys are only present when they operate
the oracle. Normally these are boys who are observing taboos
of mourning for the death of a relative. Adults also consider
that it would be very unwise to allow any boys other than these
to come near their poison because boys cannot be relied upon
to observe the taboos on meats and vegetables.
An unmarried man will seldom be present at a seance. If he
has any problems his father or uncle can act on his behalf.
Moreover, only a married householder is wealthy enough to
possess fowls and to acquire poison and has the experience to
conduct a seance properly. Senior men also say that youths are
generally engaged in some illicit love affair and would probably
pollute the poison if they came near it.
It is particularly the province of married men with house-
holds oftheir own to consult the poison oracle and no occupa-
tion gives them greater pleasure. It is not merely that they are
able to solve their personal problems; but also they are
dealing with matters of public importance, witchcraft, sorcery,
and adultery, in which their names will be associated as wit-
nesses of the oracle's decisions. A middle-aged lande is happy i
when he has some poison and a few fowls and the company '!
I·'
:iiJ
ofone or two trusted friends of his own age, and he can sit down 1~1
to a long seance to discover all about the infidelities of his wives, i
his health and the health of his children, his marriage plans, u,!
91
his hunting and agricultural prospects, the advisability of
changing his homestead, and so forth.
Poor men who do not possess poison or fowls but who are <)
compelled for one reason or another to consult the oracle will
::1
:.ij
:,1"
'Ii
,"
13 0 Oracles
persuade a kinsman, blood-brother, relative-in-Iaw, or prince's
deputy to consult it on their behalf. This is one of the main
duties of social relationships.
Control over the poison oracle by the older men gives them
great power over their juniors and is one of the main sources
of their prestige. It is possible for the older men to place the
names ofthe youths before the poison oracle and on its declara-
tions to bring accusations of adultery against them. Moreover,
a man who is not able to afford poison is not a fully independent
householder, since heis unable to initiate any important under-
taking and is dependent on the goodwill of others to inform
him about everything that concerns his health and welfare.
Women are debarred not only from operating the poison
oracle. but from having anything to do with it. They are not
expected even to speak ofit, and a man who mentions the oracle
in the presence of women uses some circumlocutory expression.
When a man is going to consult the poison oracle he says to
his wife that he is going to look at his cultivations or makes
a similar excuse. She understands well enough what he is going
to do but says nothing. Occasionally very old women of good
social position have been known to operate the poison oracle,
or at least to consult it, but such. persons are rare exceptions
and are always august persons.
The poison oracle is a male prerogative and is one of the prin-
cipal mechanisms of male control and an expression of sex an-
tagonism. For men say that women are capable of any deceit to
defy a husband and please a lover, but men at least have the
advantage that their oracle poison will reveal secret embraces.
Ifit were not for the oracle it would be oflittle 'use to pay bride-
wealth, for the most jealous watch will not prevent a woman
from committing adultery ifshe has a mind to do so. And what
woman has not? The only thing which women fear is the poison
oracle; for if they can escape the eyes of men they cannot escape
the eyes of the oracle. Hence it is said that women hate the
oracle, and that if a woman finds some of the poison in the bush
she will destroy its power by urinating on it. I once asked a
Zande why he so carefully collected the leaves used in operating
the oracle and threw them some distance away into the bush,
and he replied that it was to prevent women from finding them
and polluting them, for ifthey pollute the leaves then the poison
Oracles 131
which has been removed to its hiding-place will lose its power.
When we consider to what extent social life is regulated by
the poison oracle we shall at once appreciate how great an ad-
vantage men have over women in their ability to use it, and
how being cut off from the main means of establishing contact
with the mystical forces that so deeply affect human welfare
degrades woman's position in Zande society.
Great experience is necessary to conduct a seance in the cor-
rect manner and to know how to interpret the findings of the
oracle. One must know how many doses ofpoison to administer,
whether the oracle is working properly, in what order to take
the questions, whether to put them in a positive or negative
form, how long a fowl is to be held between the toes or in the
hand while a question is being put to the oracle, when it ought
to bejerked to stir up the poison, and when it is time to throw
it on the ground for final inspection. One must know how to
observe not only whether the fowl lives or dies, but also the
exact manner in which the poison affects it, for while it is under
the influence of the oracle its every movement is significant to
the experienced eye. Also one must know the phraseology of
address in order to put questions clearly to the oracle without
error or ambiguity, and this is no easy task when a single ques-
tion may be asked in a harangue lasting as long as five or ten
minutes. Not every man is proficient in the art, though most
adults can prepare and question the oracle if necessary. Those
who as boys have often prepared the poison for their fathers
and uncles, and who are members of families which frequent
the court and constantly consult the oracle, are the most com-
petent. Some men are very expert at questioning the oracle,
and those who wish to consult it like to be accompanied by such
a man.
VIII
Any man who is invited by the owner of the oracle poison may
attend the seance, but he will be expected to keep clear of the
oracle if he has had relations with his wife or eaten any of the ~
"
~~
prohibited foods within the last few days. It is imperative that \
the man who actually prepares the poison shall have observed 1
',!
these taboos, and for this reason the owner of the poison,
referred to in this account as the owner, generally asks a boy
··:i
 -------.- ..•.. -/
13 2 Oracles:
or man who is under tab60s of mourning to operate the oracle,
since there can be no doubt that he has kept the taboos, because
they are the same for mourning as for oracles. Such a man is
always employed when, as in a case of sudden sickness, it is
necessary to consult the oracle without warning so that there
is no time for a man to prepare himself by observation of taboos.
I shall refer to the man or boy who actually prepares the poison
and administers it to fowls as the operator. When I speak of
the questioner I refer' to the man who sits opposite to the oracle
apd addresses it and calls upon it for judgements. As he sits
a few feet from the oracle he ought also to have observed all
the taboos. It is possible for a man to be owner, operator, and
questioner at the same time by conducting the consultation of
the oracle by himself, but this rarely, if ever, occurs. Usually
there is no difficulty in obtaining the services of an operator
since a man knows which of his neighbours are observing the
taboos associated with death and vengeance. One of his com-
panions who has not eaten tabooed food or had sexual relations
with women for a day or two before the consultation acts as
questioner. Ifa man is unclean he can address the oracle from
a distance. It is better to take these precautions because contact
of an unclean person with the oracle is certain to destroy its
potency, and even the close proximity ofan unclean person may
have this result.
'The taboos which have invariably to be kept by persons who
come into contact with oracle poison are on:
Sexual relations with women.
Eating elephant's flesh.
Eating fish.
Eating mboyo vegetable (Hibiscus esculentus).
Eating morombida vegetable (Corchorus tridens).
Smoking hemp.
Some men avoid eating animals of a light colour, and such
would seem to be the rule imposed on those who come into con-
tact with a prince's oracles. Elephant's flesh and fish are for-
bidden on account of the powerful smell emitted by a man who
has eaten them. I think that it is their slimy nature that has
brought mboyo and morombida under a ritual ban. They are glu-
tinous, and when the edible parts are plucked they do not break
off cleanly but are attached to the stem by glutinous fibres
Oracles 133
which have to be drawn out. When cooked they form a sticky
mess which can be stretched like toffee. Before he comes into
contact with oracle poison, or even into close proximity to it,
a man ought to have refrained from sexual intercourse for five
or six days and to have abstained from the forbidden meats and
vegetables for three or four days. However, the length of time
during which a man ought to observe these taboos prior to
operating the oracle is not fixed, and different men give dif-
ferent estimates. Many are content to refrain from sexual inter-
course for five or even four days. If a man who has had sexual
relations is asked to operate the oracle he will say, 'I have eaten
mboyo,' and everyone will understand that he is employing a
euphemism for sexual intercourse. He may excuse himself in
similar terms if he simply does not wish to be bothered with
the work.
The owner does not pay the operator and questioner for their
services. The questioner is almost invariably either the owner
himself or one of his friends who also wishes to put questions
to the oracle and has brought fowls with him for the purpose.
It is usual to reward the operator, if he is an adult, by giving
him a fowl during the seance so that he can place one of his
own problems before the oracle. Since he is generally a man
who wears a girdle ofmourning and vengeance he will 'often ask
the oracle when the vengeance-magic is going to strike its victim.
To guard against pollution a man generally hides his poison
in the thatched roof of a hut, on the inner side, if possible, in
a hut which women do not use, but this is not essential, for a
woman does not know that there is poison hidden in the roof
and is unlikely to cbme into contact with it. The owner of the
poison must have kept the taboos if he wishes to take it down
from the roof himself, and if he is unclean he will bring the
man or boy who is to operate the oracle into the hut and indi- ~
cate to him at a distance where the poison is hidden in the
thatch. So good a hiding-place is the thatched roof of a hut :!
for a small packet of poison that it is often difficult for its owner
himselfto find it. No one may smoke hemp in a hut which lodges
oracle poison. However, there is always a danger of pollution J
and of witchcraft if the poison is kept in a homestead, and some 'I,I
men prefer to hkle it in a hole in a tree in the bush, or even :'
to build a small shelter and to lay it on the ground beneath. i
\
\,
1
I
"
"
_,i
:~
I~

134 Oracles
This shelter is far removed from human dwellings, and were a
man to come across it in the bush he would not disturb it lest
it cover some kind oflethal medicine. It is very improbable that
witchcraft will discover oracle poison hidden in the bush. I have
never seen oracle poison under a shelter in the bush, but I was
told that it is frequently housed in this manner.
Oracle poison when not in use is kept wrapped inleaves, and
at the end of a seance used poison is placed in a separate leaf-
wrapping to unused poison. The poison may be used two or
three times and sometimes fresh poison is added to it to make
it more potent. When"its action shows that it has lost its strength
they throw it away.
All good oracle poison is the same, whoever owns, operates,
and consults it. But its goodness depends on the care and virtue
of owner, operator, and consulter. As the greatest precautions
are taken with a prince's poison, it is considered more reliable
than the poison of commoners. All benge is the same material,
but people speak of ,my benge' or of 'so-and-so's benge', and they
say that the poison of one prince is absolutely reliable while
that ofanother prince is not so reliable. They make thesejudge-
ments partly on the evidence of subsequent events which prove
oracles right or wrong in their statements, and partly on the
verdicts of the king's oracle, which is the final authority. For
in the past cases would occasionally go from a provincial gov-
ernor's oracles to Gbudwe's oracle which might declare tbe m
to be in error.
IX
I will now describe the manner in which poison is
administered to fowls. The operator goes ahead of the rest of
the party in order to prepare for the test. He takes with him
a small gourdful of water. He clears a space by treading down
the grasses. Afterwards he scrapes a hole in the earth into which
he places a large leaf as a basin for the oracle poison. From
bingba grass he fashions a small brush to administer the poison,
and from leaves he makes a filter to pour the liquid poison into
the beaks of the fowls; and from other leaves he makes a cup
to transfer water from the gourd to the poison when it needs
to be moistened. Finally, he tears off some branches of nearby
shrubs and extracts their bast to be used as cord for attaching
Oracles 135
to the legs of fowls which have survived the test so that they
can be easily retrieved from the grass when the business of the
day is finished. The operator does not moisten the poison till
the rest of the party arrive.
There may be only one man or there may be several who
have questions to put to the oracle. Each brings his fowls with
him in an open-wove basket. As it has been agreed beforehand
where the oracle consultation is to take place they know where
to foregather. As each person arrives he hands over his basket
offowls to the operator who places it on the ground near him.
A man who is used to acting as questioner sits opposite to it,
a few feet away ifhe has observed the taboos, but several yards
away if he has not observed them. Other men who have not
kept the taboos remain at a greater distance.
When everyone is seated they discuss in low tones whose fowl
they will take first and how the question shall be framed. Mean-
while the operator pours some water from the gourd at his side
into his leaf cup and from the cup on to the poison, which then
effervesces. He mixes the poison and water with his finger-tips
into a paste of the right consistency and, when ins.tructed by
the questioner, takes one of the fowls and draws down its wings
over its legs and pins them between and under his toes. He takes
his grass brush, twirls it round in the poison, and folds it in
the leaf filter. He holds open the beak of the fowl and tips the
end of the filter into it ~nd squeezes the filter so that the liquid
runs out of the paste into the throat of the fowl. He bobs the
head ofthe fowl up and down to compel it to swallow the poison.
At this point the questioner, having previously been in-
str'Ucted by the owner of the fowl on the facts which he is to
put before the oracle, commences to address the poison inside
the fowl. He continues to address it for about a couple of
minutes, when a second dose of poison is usually administered.
If it is a very small chicken two doses will suffice, but a larger
fowl will receive three doses, and I have known a fowl receive
a fourth dose, but never more than four. The questioner does
not cease his address to the oracle, but puts his questions again
and again in different forms, though always with the same
refrain, 'If such is the case, poison oracle kill the fowl,' or 'If
such is the case, poison oracle spare the fowl.' From time to
time he interrupts his flow of oratory to give a technical order
 136 Oracles
"3
to the operator. He may tell him to give the fowl another dose
of poison or to jerk it between his toes by raising and lowering
his foot (this stirs up the poison inside the fowl). When the last
dose of poison has been administered and he has further
addressed it, he tells the operator to raise the fowl. The operator
takes it in his hand and, holding its legs between his fingers
so that it faces him, gives it an occasional jerk backwards and
forwards. The questioner redoubles his oratory as though the
verdict depended upon his forensic efforts, and if the fowl is
not already dead he then, after a further bout of oratory, tells
the operator to put it on the ground. He continues to address
the poison inside the fowl while they watch its movements on
the ground.
The poison affects fowls in many ways. Occasionally it kills
them immediately after the first dose, while they are still on
the ground. This seldom happens, for normally a Jowl is not
seriously affected till it is removed from the ground and jerked
backwards and forwards in the hand. Then, ifit is going to die,
it goes through spasmodic stretchings of the body and closing
of the wings and vomits. After several such spasms it vomits
and expires in a final seizure. Some fowls appear quite un-
affected by the poison, and when, after beingjerked backwards
and forwards for a while, they are flung to the ground peck
about unconcernedly. Those fowls which are unaffected by the
poison generally excrete as soon as they are put to earth. Some
fowls appear little affected by the poison till put to earth, when
they suddenly collapse and die. It is very seldom that a fowl'
seriously affected by the poison finally recovers.
One generally knows what the verdict is going to be after
the fowl has been held in the hand for a couple of minutes. H
itappears certain to recover the operator ties bast to its leg and
throws it to the ground. Hit appears certain to die he does not
trouble to tie bast to its leg, but lays it on the earth to die. Often
when a fowl has died they draw its corpse in a semicircle round
the poison to show it to the poison. They then cut off a wing
to use as evidence and cover the body with grass. Those fowls
which survive are taken home and let loose. A fowl is never
used twice on the same day.
There is no stereotyped speech-no formula-in which the
oracle must be addressed. Nevertheless, there are traditional
Oracles 137
refrains, pieces of imagery, compliments to the oracle, ways of
formulating a question, and so forth which Occur in every con-
sultation.
The main duty of the questioner is to see that the oracle fully
understands the question put to it and is acquainted with all
facts relevant to the problem it is asked to solve. They address
it with all the care for detail that one observes in court cases
before a prince. This means beginning a long way back and
noting over a considerable period of time every detail which
might elucidate the case, linking up facts into a consistent
picture of events, and the marshalling of arguments into a logi-
cal and closely knit web of sequences and interrelations offact
and inference. Also the questioner is careful to ~ention to the
oracle again and again the name of the man who is consulting
it, and he points him out to the oracle with his outstretched
arm. He mentions also the name ofhis father, perhaps the name
of his clan, and the name of the place where he resides, and
he gives similar details of other people mentioned in the
address.
An address consists usually of alternate directions. The first
sentences outline the question in terms demanding an affirma-
tive answer and end with the command, 'Poison oracle kill the
fowl.' The next sentences outline the question in terms demand_
ing a negative answer and end with the command, 'Poison
oracle spare the fowl.' The consulter then takes up the question
again in terms asking an affirmative answer; and so on. If a
bystander considers that a relevant point has been left out he ,I
interrupts the questioner, who then makes this point.
The questioner has a switch in his hand, and while question-
ing the oracle beats the ground, as he sits cross-legged, in front
ofit. He continues to beat the ground till the end of his address.
Often he will gesticulate as he makes his points, in the same
manner as a man making a case in court. He sometimes plucks
grass and shows it to the poison and, after explaining that there
is something he does not wish it to consider, throws it behind
him. Thus he tells the oracle that he does not wish it to consider
the question of witchcraft but only of sorcery. Witchcraft is
wingi, something irrelevant, and he casts it behind him. The
imagery used is specially noteworthy. It is seldom that the
. oracle is addressed without analogies and circumlocutions.
138 Oracles
Thus in asking whether a man has committed adultery one
frames the question in some such manner as follows:
·Poison oracle, poison oracle, you are in the throat of the fowl. That
man his navel joined her navel; they pressed together; he knew her
as woman and she knew him as man. She has drawn badiabe (a leaf
used as a towel) and water to his side (for ablutions after intercourse);
poison oracle hear it, kill the fowl.
While the fowl is undergoing its ordeal men are attentive to
their behaviour. A man must tighten and spread out his bark-
cloth loin-covering lest he expose his genitals, as when he is sit-
ting in the presence of a prince or parent-in-law. Men speak
in a low voice as they do in the presence of superiors. Indeed,
all conversation is avoided unless it directly concerns the pro-
cedure ofconsultation. Ifanyone desires to leave before the pro-
ceedings are finished he takes a leaf and spits on it and places
it where he has been sittng. I have seen a man who rose for
a few moml:'nts only to catch a fowl which had escaped from
its basket place a blade of grass on the stone upon which he
had been sitting. Spears must be laid on the ground and not
planted upright in the presence of the poison oracle. Azande
are very serious during a seance, for they are asking questions
of vital importance to their lives and happiness.
x
Basically, the system ofquestion and answer in oracle consulta-
tions is simple. There are two tests, the bambata sima, or first
test, and the gingo, or second test. If a fowl dies in the first test
then another fowl must survive the .second test, and if a fowl
survives the first test another fowl must die in the second test
for the judgement to be accepted as valid. Generally the ques-
tion is so framed that the oracle will have to kill a fowl in the
first test and spare another fowl in the corroborative test to give
an affirmative reply, and to spare a fowl in the first test and
kill another fowl in the corroborative test to give a negative
reply; but this is not invariably the case, and questions are
sometimes framed in an opposite manner. The killing of a fowl
does not give in itself a positive or negative answer. That de-
pends upon the form of the question. I will illustrate the usual
procedure by an example:
Oracles 139
A.
First Test. lfX has committed adultery poison oracle kill the fowl.
If X is innocent poison oracle spare the fowl. The fowl dies.
Second Test. The poison oracle has declared X guilty of adultery
by slaying the fowl. Ifits declaration is true let it spare this second
fowl. The fowl survives.
Result. A valid verdict. X is guilty.
B.
First Test. If X has committed adultery poison oracle kill the fowl.
If X is innocent poison oracle spare the fowl. The fowl lives.
Second Test. The poison oracle has declared X innocent of adultery
by sparing the fowl. Hits declaration is true let it slay the second
fowl. The fowl dies..
Result. A valid verdict. X is innocent.
c.
First Test. If X has committed adultery poison oracle kill the fowl.
If X is innocent poison oracle spare the fowl. The fowl dies.
Second Test. The poison oracle- has declared X guilty of adultery
by slaying the fowl. Hits declaration is true let it spare the second
fowl. The fowl dies.
Result. The verdict is contradictory and therefore invalid.
D.
First Test. If X has committed adultery poison oracle kill the fowl.
If X is innocent poison oracle spare the fowl. The fowl survives.
Second Test. The poison oracle has declared X innocent of adultery
by sparing the fowl. Ifits declaration is true let it slay the second
fowl. The fowl survives.
Result. The verdict is contradictory and therefore invalid.
In the two tests one fowl must die and the other must live i
if the verdict is to be accepted as valid. If both live or both I
die the verdict is invalid and the oracle must be consulted on ,I
I.~~
the matter a second time on another occasion. If the supply
of oracle poison is sufficient the two tests may be made during .
the same seance, especially when the matter is important and
1
urgent. Very often, however, a test is not completed at a single
seance, as will be observed in the tables that follow, for one
of these reasons:
(I) The other part of the test may have been carried out pre-
viously or may be carried out at a future seance. Sometimes
140 Oracles
a long interval elapses between two tests because the first one
is considered sufficient justification for commencing an under-
taking, but a second test has to be made before the undertaking
is far advanced, e.g. a man is betrothed to a girl and begins
to pay bride-spears to her father on the authority of a single
test and leaves the corroborative test till months later. But the
girl will not come to live with him permanently till both tests
have been made. (2) One of the lesser oracles may have been
consulted earlier so that a single verdict of the poison oracle
is therefore regarded as an oracular confirmation. (3) Often
Azande consider a single test sufficient, especially if the oracle
gives its answer decisively by killing the fowl without hesitation.
They are able to economize their oracle poison by this means.
(4) Many confirmations ofverdicts are contained in the oracle's
answers to other questions, e.g. a man asks whether a witch
will die if a certain kinsman observes taboos of vengeance-
magic. The oracle says 'Yes' . He then asks whether the kinsman
will die during the period he is under taboos. If the oracle
says 'No' it confirms its previous verdict because the life of the
kinsman is bound up with the accomplishment of vengeance.
(5) Sometimes a single fowl is used to confirm different ques-
tions. If in answer to two different questions the oracle killed
two fowls it may then be asked to spare a third fowl to confirm
both its verdicts at the same time. (6) When a serious matter
is not at stake Azande are sometimes content merely to know
that the oracle is functioning correctly, and being assured of
this, are prepared to accept its single statements and to dispense
with repetitions ofjudgement. Thus five unconnected questions
may be asked in a seance. The oracle spares fowls in answer
to the first four questions and then kills a fowl in answer to the
fifth question. This shows that the action bf the particular
bundle of poison is discriminating and therefore its first four
verdicts may be assumed to be valid.
But two tests are essential in any question that concerns the
relations between two persons, especially when they involve
legal issues.
XI
The following consultations of the poison oracle are given to
show the type of questions asked and the order of asking, and
Oracles 14 1
to enable the reader to judge for himself the proportion offowls
that die, the number of doses of poison they receive, and the
order of deaths and survivals. I was present at both the seances
recorded, and many ofthe questions concern persons connected
with my household and their relatives.
SEANCE I
(I) Should X take on the taboos of mourning and vengeance for the
death of Magadi till vengeance be accomplished? The fowl DIES,
giving the answer 'Yes'.
(2) If X takes on the taboos of mourning for Magadi will he die in
consequence (i.e. if, through carelessness in its use, the magical
medicine he has sent out against Magadi's murderer should turn
back upon X himself? This would also be a corroboration of the
first question, since if X were to die then vengeance would not
be accomplished during his period of mourning.) The fowl DIES,
giving the answer 'Yes'. (These two verdicts contradicted one
another and a short discussion followed. One man present said
that since Magadi died of leprosy his death ought not to be
avenged, and that for this reason the oracle had given contradic-
tory verdicts. This opinion was rejected by others.)
(3) IfAdiyambio, who is suffering from a deep-seated ulcer, remains
in our government settlement, will he die? The fowl SURVIVES,
giving the answer 'No'.
(4) IfBamina lives in the new homestead which he has just built for
himself will he die? The fowl DIES, giving the answer 'Yes'.
(5) If Bamina remains in his old homestead will he die? The fowl
DIES, giving- the answer 'Yes'.
(6) IfBamlna goes to live in the government settlement of Ndoruma
will he die? The fowl SURVIVES, giving the answer 'No'.
(7) (Corroboration of the last question.) Did the oracle speak truly
when it said that Bamina would not die ifhe went to live in the
government settlement of Ndoruma? The fowl SURVIVES, giving
the answer 'No'. (The answers to questions 6 and 7 therefore con-
tradicted one another. Someone suggested that the oracle was
tired like a chief who has been sitting for hours listening to cases
i·,j
in his court and is weary. Another man said that the oracle saw ~::r
some misfortune ahead, which was not death yet was a serious
misfortune, and had taken this way of warning Barnina. In any
case, the verdicts taken together were considered a bad augury
and there was a short discussion about who was threatening the
welfare of Bamina. Mbira gave it as his opinion that the danger
14 2 Oracles
was from sorcery and not from witchcraft since witchcraft does
not pursue a man from one place to another in this manner but
ceases to trouble him if he leaves his homestead and goes to live
elsewhere. )
(8) They now ask the oracle about two men, one called Pilipili and
the other a man of -the Bangombi clan who had once married
Bamina's daughter but whose bride-spears had been returned to
him. Are either of these two men threatening Bamina with witch-
craft or with bad magic? The fowl DlES, giving the answer 'Yes'.
The seance had to be closed at this point as there was not
enough poison left to continue consultations.
SEANCE II
(I) Since by an earlier consultation it has been determined that the
daughter ofMamenzi, the wife ofMekana, is in a bad 'condition',
is the evil influence that hangs over her from the homestead of
Mekana or from the homestead of her paternal grandfather (who
had been given her bride-spears by her father as 'first-fruits')?
Ifit is from Mekana's homestead, poison oracle spare the fowl.
If it is from her grandfather's homestead, poison oracle kill the
fowl. (It may be remarked that this is a very unusual way ofput-
ting a question to the oracle since it does not allow for a third
alternative: that the witch is a member of some household other
than the two mentioned. The procedure might even be regarded
as incorrect. However, the husband was so certain that the evil
influence which threatened his wife could only have arisen from
jealousy in his own household, or from displeasure in the house-
hold of his parents-in-law, that the question appeared to him
legitimate. Moreoever, it was always possible for the oracle to
show that neither household was responsible by killing or sparing
both fowls in the double test, or even by the way in which it
affected the fowls during the tests.) The fowl DIES, saying that
the evil influence is from the homestead of the girl's grandfather.
(One dose of the poison was administered.)
(2) The rubbing-board oracle has said that a man named Sueyo
made the magic which caused Kisanga such violent sickness. The
question is now asked, 'Is the statement of the rubbing-board cor-
rect? Ifso, poison oracle kill the fowl!' The fowl SURVIVES, giving
the answer 'No'. (Two doses administered.)
(3) X's mother lies seriously ill. Is her sickness due to Basa? If so,
poison oracle kill the fowl. IfBasa is not responsible, poison oracle
spare the fowl. The fowl SURVIVES, giving the answer 'No'. (Two
doses administered.)
Oracles 143
(4) (Corroborative verdict to question No. (I).) If the evil influence
that threatens his wife is due to Mekana's household, then poison
oracle kill the fowl. If the evil influence emanates from the wives
of his wife's grandfather, then poison oracle spare the fowl. The
fowl SURVIVES, confirming that evil influence is from the home-
stead 'of the girl's grandfather. (Two doses administered.)
(Mekana afterwards approached his father-in-law so that the
womenfolk ofhis household might all collect and blowout water
in sign ofgoodwill. He did not venture to single out any particular
'mother-in-law'.)
(5) Since the oracle (test No. (3)) said that the sickness ofX's mother
is not due to Basa, X now asks whether it is due to the wives of
Y. If the wives of Yare responsible, poison oracle kill the fowl.
The fowl DIES, giving the answer 'Yes'. (One dose administered.)
(6) (We now return to question No. (2).) It having been determined
that Sueyo was not responsible for Kisanga's sickness, he asks
whether the sorcerer lives on our side of the new part of the
government settlement? If he lives there, poison oracle kill the
fowl. The fowl SURVIVES, giving the answer 'No'. (Two doses
administered.) (This verdict, combined with three previous ver-
dicts on the matter, proved that the sorcerer did not live anywhere
in Our settlement.)
(7) (We return to the subject of Mekana's wife already dealt with
in questions (I) and (4)·) If there is any\?ne else besides the wives
of his wife's grandfather who threatens her health, or if after the
fowl's wing has been presented to them to blow water on to it
they will still exercise an evil influence over her, then poison
oracle kill the fowl. If, on the other hand, there is no one else
to fear besides the wives of his wife's grandfather, and if they will
blowout water on to the fowl's wing with sincerity and withdraw
their evil influence, then poison oracle spare the fowl. The fowl "'!
SURVIVES, indicating that there will be nothing more to fear.. (Two
doses administered.)
(8) (We return to the question of X's mother already dealt with in
tests Nos. (3) and (5)·) It having been determined that the wives
of Yare responsible for the sickness of his mother, X now asks
whether they are alone responsible or whether Y himself has
encouraged and assisted them in bewitching the old woman. If
Y is guilty, then poison oracle kill the fowl. IfY is innocent, then
poison oracle spare the fowl. The fowl DIES, saying that Y is re-
sponsible. (One dose administered.)
This second seance provides an example of a Wholly success-
ful consultation of the oracle. I would call attention to the
144 Oracles
Oracles 145
manner in which an assortment ofquestions is arranged. There
rives in important cases involving princes. The address was
are three problems to be solved, and there are eight fowls by
made in the same idiom as an address to fowls. The poison was
which to solve them. The questions concern the welfare of
mixed with water in a gourd. The boy, seated on the ground
Mekana's wife, the health ofa woman referred to as X's mother,
and wearing a girdle of bingba grass, drank the poison, and then
and the identification of the sorcerer who has caused Kisanga
the questioner shook hand-bells and addressed the poison inside
such grievous sickness. When several persons have questions to
him. When he had finished his address he rubbed the gour:d
put before the oracle one does not thrash out one problem and
on the boy's head and ordered him to rise. If the boy had
then turn to the next, but generally, as on this occasion, each
reached the fowl's wing and returned with it they would again
person is allowed to ask a question in turn. In the second round
have addressed the poison within him and would then have told
each person tries to procure corroborative verdicts or asks subsi-
him to replace the fowl's wing. They would afterwards have
diary questions. If one man has more fowls than the others he
uttered a third and final address and told the lad to fetch the
is able to ask more questions, but he allows others to place their wing again. The test would then have ended.
problems in between his queries. This is not simply a matter
If the poison were going to kill a boy it would not kill him
of courtesy but also rests on a notion that after a problem has
while he sat still on the ground, though he would suffer spasms
been put to the oracle and it has given its answer it should be
ofpain that would make him stretch his arms backwards, gasp-
granted time to turn the matter over at leisure before it corro-
ing for breatho When a boy fell to the ground efforts were made,
borates its first answer and gives a final verdict. The poison used
with the king's consent, to revive him by administering a slimy
at this seance was at once seen to be discriminating. It killed
mixture made from the mbo?fo plant, the kpo?fo tree, and salt.
the first fowl and showed that it was not impotent because when
This made him vomit the poison. Afterwards they carried him
benge is impotent all the fowls survive. It spared the second fowl,
to a brook-side and laid him in the shade and poured cold water
showing that it was not stupid, over-potent poison, for when over his face.
it is such all the fowls die. It spared several other fowls, but
at the finish killed the last fowl, showing that it maintained its
potency. Azande look to these evidences in every test to estab-
lish that the poison is good.
XII

It remains to give an account of how human beings used to

drink oracle poison in the old days. Some care is necessary in
taking account of the Zande phrase mo mbiri benge, 'You drink
oracle poison,' because this is a usual expression of a prince
when he means no more than, 'You must submit your case to
the poison oracle.' But in the past people sometimes, though
very rarely, actually drank. poison themselves. This might
happen in two ways. A man accused of some serious offence
might offer to drink poison after an oracular test with fowls had
gone against him. Likewise, if a woman accused a man ofhav-
ing committed adultery with her he could demand that both
he and the woman should drink poison.
Oracle poison was also occasionally administered to boy cap-

CHAPTER IX
Problems arising from consultation
of the Poison Oracle
I HAVE described to many people in England the facts related
in the last chapter and they have been, in the main, incredulous
or contemptuous. In their questions to me they have sought
to explain away Zande behaviour by rationalizing it, that is
to say, by interpreting it in terms of our culture. They assume
that Azande must understand the qualities of poisons as we
understand them; or that they attribute a personality to the
oracle, a mind that judges as men judge, but with higher
prescience; or th;;l.t the oracle is manipulated by the operator
whose cunning conserves the faith of laymen. They ask what
happens when the result ofone test contradicts the other which
it ought to confirm if the verdict be valid; what happens when
the findings of oracles are belied by experience; and what
happens when two oracles give contrary answers to the same
question.
These same, and other, problems, naturally occurred to me
in Zandeland, and I made inquiries into, and observations on, '
i<
those points which struck me as being important;--and in the
:i present section I record my conclusions. Before setting them
down I must warn the reader that we are trying to analyse beha-
viour rather than belief. Azande have little theory about their
oracles and do not feel the need for doctrines.
I have translated the word benge as 'poison creeper', 'oracle
poison', and 'poison oracle', in accordance with the context.
But it is necessary to point out that Zande ideas about benge
are very different from notions about poisons prevalent among
the educated classes of Europe. To us it is a poison, but not
! to them.
It is true that benge is derived from a wild forest creeper and
that its properties might be supp~sed to reside in the creeper,
:t
Oracles 147
i.e. to be natural properties, but in Zande eyes it only becomes
the benge of oracle consultations (and they have no interest in
it outside this situation) when it has been prepared subject to
taboos and is employed in the traditional manner. Properly
speaking, it is only this manufactured benge which is benge at
all in Zande opinion. Hence Azande say that if it is deprived
of its potency for some reason or other it is 'just an ordinary
thing, mere wood'.
Therefore, to ask Azande what would happen if they were
to administer oracle poison to a fowl without delivering an
address or, ifthey were to administer an extra portion of poison
to a fowl which has recovered from the usual doses, or, if they
were to place some of the poison in a man's food, is to ask silly
questions. The Zande does not know what would happen, he
is not interested in what would happen, and no one has ever
been fool enough to waste good oracle poison in making such
,\ pointless experiments. Proper benge is endowed with potency
~ by man's abstinence and his knowledge of tradition and will
only function in the conditions of a seance.
When I asked a Zande what would happen if you went on
administering dose after dose of poison to a fowl during a con- j
sultation in which the oracle ought to spare the fowl to give
'I
J
the right answer to the question placed before it, he J;"eplied that I
he did not know exactly what would happen, but that he sup-
posed sooner or later it would burst. He would not countenance ")
>!I.'"
the suggestion that the extra poison would otherwise kill the ,
fowl unless the question were suddenly reversed so that the
oracle ought to kill the fowl to give a correct answer when, of
course, it would at once die.
I t is certain that Azande do not regard the reactions of fowls
to benge and the action of benge on fowls as a natural process,
that is to say, a process conditioned only by physical causes.
The oracle is not to them a matter of chance, like the spinning
of a coin, by which they are agreed to abide. Indeed, we may
ask whether they have any notion that approximates to what
we mean when we speak of physical causes.
Yet it might still be possible for Azande to have a crude
common-sense notion ofpoisons. They might know that there are
certain vegetable products that will kill men and beasts, with-
out attributing supra-sensible properties to them. Certainly
148 Oracles
Europeans often attribute knowledge of poisons to Azande and
to other peoples of the Southern Sudan. No evidence for the
homicidal use of poison has yet been produced, nor is likely to
be. If there is one product possessed by Azande that is certainly
poisonous it is benge, and daily its lethal properties are demon-
stratedon fowls, and sometimes have been demonstrated on men,
yet they have no idea that it might be possible to kill people by
adding it to their food. Though men are frequently suspected
of using one kind or other of bad medicines to slay their neigh-
bours, no one has ever conceived of a man using benge as a
means of murder, and if you suggest it to him a Zande will tell
you that benge would not be any good for the purpose.
Yet it is not always easy to reconcile Zande doctrines with
their behaviour and with one another. They say that men will
sometimes eat fowls after having cleansed them of poison, and
this action would imply a knowledge of the natural properties
of benge that they refuse to allow in other situations. The owner
of a dead fowl may have its stomach and neck removed and
the fowl prepared for food. My informants sai,d that they try
to remove all the poison from the carcass. Probably the practice
is rare, since as a rule chickens are used which are too small
for culinary purposes. Generally the fowls are thrown away or
placed in a tree for birds to devour after their wings have been
cut off. Moreover, a young man would not eat fowls killed by
the oracle, so that the statement applies only to old men, and
perhaps only to those who are not very particular about their
food. When I protested against the statement that persons eat'
poisoned fowls I was asked, 'What harm can it do a man since
no one addresses it?' Mekana once remarked to me that it
would be rather ajoke to address the oracle poison in the belly
ofan elder who had eaten a fowl which had died in an oracle
test. One might say, he suggested, 'If so-and-so (naming the
elder) slept with his wife last night, poison oracle kill him.' I
think that Mekana was hardly serious in this suggestion.
Nevertheless, the very fact of cleansing fowls of poison suggests
that Azande are to some extent aware of its natural properties.
Some Azande hold that the poison will deteriorate with age,
and all are aware that so'me poison is stronger than others and
that it becomes more potent when exposed to the sun and less
potent when diluted in water. They know that if a dog eats
Oracles 149
a fowl that has succumbed to the oracle it may die (it is possible
that they conceive of the oracle still working inside the dog and
answering the question put to it earlier, but I have no evidence
that this is the case. It may also be possible that when
men cleanse fowls killed by the oraCle before eating them
they are afraid lest the poison go on answering the question
inside them and kill them. I have no doubt that a Zande L
(:'!
might give so characteristically mystical a reason for his be-
haviour.) :1'I
ij
II Ij
:1
Without laboratory experiments it is impossible to see any uni- '~
"'::1
formitiesin the working of the oracle. Bare observation by itself ,!j'Hl
is insufficient to explain why some fOWls die and others survive. ,I,
As a matter of fact, Azande act very much as we would act
in like circumstances and they make the same kind of observa-
tions as we would make. They recognize that some poison is :1'
strong and other poison is weak and give more or fewer doses
according to the kind they are using. One often hears it said
"I:I, ~
I
during a seance, 'It is not strong enough,' 'You have given the jl
"'~
fowl enough,' and like expressions. But Azande are dominated if'!l
I;~'I
'
by an overwhelming faith which prevents them from making :l
experiments, from generalizing contradictions between tests, I,''i )
between verdicts of different oracles, and between all the
oracles and experience. To understand why it is that Azande
do not draw from their observations the conclusions we would
draw from the same evidence, we must realize that their atten-
I
I'I"
W
":;!
tion is fixed on the mystical properties of the poison oracle and
that its natural properties are of so little interest to them that
they simply do not bother to consider them. To them the
creeper is something other than the final product of manu-
facture used in ritual conditions, and the creeper scarcely enters
into their notions about the oracle. If a Zande's mind were not
fixed on the mystical qualities of benge and entirely absorbed
by them he would perceive the significance of the knowledge
he already possesses. As it is the contradiction between his
beliefs and his observations only become a generalized and glar-
ing contradiction when they are recorded side by side in the
pages of an ethnographic treatise. But in real life these bits of
knowledge do not form part of an indivisible concept, so that
15 0 Oracles
when a man thinks of benge he must think of all the details I
have recorded here. They are functions of different situations
and are uncoordinated. Hence the contradictions so apparent
to us do not strike a Zande. Ifhe is conscious of a contradiction
it is a particular one which he can easily explain in terms of
his own beliefs.
It is evident that the oracle system would be pointless if the
possibility of benge being a natural poison, as an educated Euro-
pean would regard it, were not excluded. When I used at one
time to question Zande faith in their poison oracle I was met
sometimes by point-blank assertions, sometimes by one of the
evasive secondary elaborations of belief that provide for any
particular situation provoking scepticism, sometimes by polite
pity, but always by an entanglement oflinguistic obstacles, for
one cannot well express in its language objections not formu-
-lated by -3. culture.
Azande observe the action ofthe poison oracle as we observe
it, but their observations are always subordinated to their
beliefs and are incorporated into their beliefs and made to
explain them and justify them. Let the reader consider any
argument that would utterly demolish all Zande claims for the
power of the oracle. If it were translated into Zande modes of
thought it would serve to support their entire structure of belief.
For their mystical notions are eminently coherent, being inter-
related by a network of logical ties, and are so ordered that
they never too crudely contradict sensory experience but, in-
stead, experience seems to justify them. The Zande is immersed
in a sea of mystical notions, and if he speaks about his poison
oracle he must speak in a mystical idiom.
Ifwe cannot account for Zande faith in their poison oracle
by assuming that they are aware that it is a poison and are wil-
ling to abide by the chance of its action on different fowls we
might seek to comprehend it by supposing that they personify
it. Given a mind, the Zande oracle is not much more difficult
to understand than the Delphic Oracle. But they do not per-
sonify it. For, though it would seem to us that they must regard
the oracles as personal beings, since they address them directly;
in fact the question appears absurd when framed in the Zande
tongue. A boro, a person, has two hands and two teet, a head,
a belly, and so on, and the poison oracle has none ofthese things.
Oracles 15 1
It is not alive, it does not breathe or move about. It is a thing.
Azande have no theory about it; they do not know why it works,
but only that it does work. Oracles have always existed and
have always worked as they work now because such is their ~1
nature.
If you press a Zande to explain how the poison oracle can
see far-off things he will say that its mbisimo, its soul, sees them. ,I'~i.i
It might be urged that if the poison oracle has a soul it must {i
be animate. Here we are up against the difficulty that always
arises when a native word is translated by an English word.
I have translated the Zande word mbisimo as 'soul' because the
i.
·'..I I
d
notion this word expresses in our own culture is nearer to the
13
~
Zande notion of mbisimo ofpersons than any other English word.
The concepts are not identical, and when in each language-the ~
I
word is used in a number ofextended senses it is no longer poss- ~
ij
ible to use the original expressions in translation without risk ~
of confusion and gross distortion. In saying that the poison ~
oracle has a mbisimo Zande mean little more than 'it does some-
i
I~
thing' or, as we would say, 'it is dynamic'. You ask them how
it works and they reply, 'It has a soul.' If you were to ask them ,1'1
how they know it has a 'soul', they would reply that they know I'I,:II.
because it works. They are explaining mystical action by nam- :II:i~'
ing it. The word mbisimo describes and explains all action of 'If
a mystical order.
It becomes quite evident that Azande do not regard oracles
as persons when we consider the rubbing-board oracle and the
termites oracle. The rubbing-board is an instrument made by'
man out of wood and it only becomes an oracle when treated,
and afterwards operated, in a certain manner, and if a taboo :\
is broken it becomes once again merely shapen wood without
power to see the future. Termites are certainly not corporeally
or psychically persons. They are simply termites and nothing
more, but if they are approached in the correct manner they
are endowed with mystical powers.
It is difficult for us to understand how poison, rubbing-board,
termites, and three sticks can be merely things and insects and
yet hear what is said to them and foresee the future and reveal
l'
the present and past, but when used in ritual situations they
cease to be mere things and mere insects and become mystical
agents. And, since oracles are endowed with their nowers by
1,,'1'
152 Oracles
man himself, through man they may lose those powers. If a
taboo is broken they become once again mere things, insects,
and bits of wood.
III
I t will at once occur to a European mind that a likely reason
why one fowl dies and another lives is because more or larger
doses of poison are administered to the one than to the other,
and he is likely to jump to the conclusion that the verdict de-
pends on the skill of the operator. Indeed, a European is prone
to assume that the operator cheats, but I believe that he is
wrong in this assumption. It is true that the number and size
of doses given to fowls varies, and that even fowls of the same
size do not always receive the same number of doses. But
to suppose that Azande cheat is entirely to misunderstand their
mentality. What would be the object in cheating? Today the
declarations of the poison oracle are no longer recognized as
evidence of murder or adultery, so that it can no longer be
used as an instrument ofjustice and profit, and the usual ques-
tions placed before it concern the health and welfare ofthe ques-
tioner and his family. He wants to know always whether witch-
craft is threatening his interests and, ifso, who is the witch who
has doomed him to some ill fate. Cheating, far from helping
him, would destroy him, for instead of being able to approach
the right witch and thus be released from his doom, he will
approach the wrong person, or no person at all, and fall an
inevitable victim to the fate that awaits him. It is entirely
against his interests that trickery be used. It would probably
result in his death. Even in questions ofmarriage where it might
seem to the advantage ofa Zande to obtain a favourable verdict
inorder that he might marry a certain girl, it would in fact
be fatal to cheat, because were he to obtain an inaccurate ver-
dict it would merely mean that his wife would die shortly after
marriage.
It might, however, be urged that the consulter of the oracle
is one person and the operator another, and that the feelings
and purpose of the consulter are of less account than the cun-
ning of the operator. This, as we shall see in the next chapter,
may be a fair comment on the working of the rubbing-board
oracle, but it is not apposite to the poison oracle for the follow-
Oracles 153
ing reasons: (I) The operator performs in public. His audience,
all parties interested in the dispute or inquiry, sit a few feet away
and can see what he does, and they largely direct his actions.
(2) It was evident to me on the many occasions I witnessed
consultations that the operator was just as little aware of what
the result of a test was going to be as I or any of the other
observers were. I judged from his actions, speech, and expres-
sion that he regarded himself as a mechanical server to, and
in no way a director of, the oracle. (3) Sometimes the consulter
of the oracle is the operator of it. A man who believes what
Azande believe about witchcraft and oracles and then cheats
would be a lunatic. (4) I have witnessed cases when it has been
to the interests of the operator that the fowls shall live and they
have died, and vice versa. (5) There is no special class of opera-
tors. They are not a corporation or closed association. Most
II adult males know how to operate the oracle, and anyone who
r wishes to operate it may do so. You cannot deceive one who
practises your particular brand of deception. (6) The operators
are generally boys of between 12 and 16, old enough to know
and keep food taboos and young enough to be able to refrain
from sexual intercourse. These innocents are the most unlikely
people in Zandeland to know how to cheat, and are besides
unconcerned, as a rule, with the adult problems that are pre- :~J
"
sented to the oracle. (7) As often as not when there are two '~
, 1'1
tests about a question the oracle contradicts itself (8) Azande i!
';~
"'.1'
do not understand that benge is a natural poison and therefore :!tl
d
do not know that trickery of this kind would even be possible. .~
They will say of the rubbing-board oracle that a man has
cheated with it, but one never hears it suggested that a man
might have unfairly manipulated the poison oracle. .~i~i
l!1
The difference in the number of doses given.to fowls is due 111
to certain technical rules in operafing the oracle. There are a
usual number of doses for fowls ofdifferent sizes, but the oracle
gives its answers through the fowls, this being the only way in
which it can speak; so that it is convenient that the fowl shall
be seen to be affected by the poison, for then they know that
it has heard the question, has considered it, and is replying to
it. Therefore, if after two doses the fowl does not seem to be
at all affected, even though this is the usual number of doses
for a fowl of that size, they may give it a third dose. If the fowl
154 Oracles
is still unaffected they know that the oracle is going to give a
clear verdict by sparing it and has answered without qualifica-
tions, for since it has killed other fowls on the same day it is
known to be good benge which can kill fowls if it wishes to do
so.
I have observed that Azande sometimes give fewer doses in
the second test, the gingo, than in the first test. They are not
trying to cheat but do not want to waste valuable poison. The
purpose of the second test is to ascertain that the oracle was
functioning correctly when it gave its first answer. It can
show this after one or two doses as clearly as after three or four
doses, and it is merely waste of good poison to give the extra
doses.
Azande realize that in civil disputes, concerning witchcraft
or adultery for example, it is possible for the man chosen to
consult the poison oracle about the point at issue to cheat in
another way. A man would not tamper with the poison because
he does not believe it possible to alter the verdict of an oracle
once the poison has been administered to a fowl, but he can
produce a hen's wing without ever consulting the poison oracle
at all, for he may merely kill a fowl and cut off its wing. Azande
say that this sometimes happens, but that the danger of its
occurrence is smaU because the elder who makes the test norm-
ally takes two or three witnesses with him. Moreover, it is poss-
ible for a man who is convinced that he 'has not been given
a fair test to appeal to the king, and ifhis poison oracle declares
the man to be innocent the king will send for the elder and
tell him that he is a cheat and a liar and may never again con-
duct official consultations.
IV
What explanation do Azande offer when the oracle contradicts
itself? Since Azande do not understanp the natural properties
of the poison they cannot explain the contradiction scientific-
ally; since they do not attribute personality to the oracle they
cannot account for its contradictions by volition; and since they
do not cheat they cannot manipulate the oracle to avoid con-
tradictions. The oracle seems so ordered to provide a maximum
'number ofevident contradictions for, as we have seen, in impor-
tant issues a single test is inacceptable and the oracle must slay
Oracles 155
one fowl and spare another if it is to deliver a vali'd verdict.
As we may well imagine, the oracle frequently kills both fowls
or spares both fowls, and this would prove to us the futility of
the whole proceeding. But it proves the opposite to Azande.
They are not surprised at contradictions; they expect them.
Paradox though it be, the errors as well as the valid judgements
of the oracle prove to them its infallibility. The fact that the
oracle is wrong when it is interfered with by some mystical
i,
"I
power shows how accurate are its judgements when these
powers are excluded.
The secondary elaborations of belief that explain the failure
of the oracle attribute its failure to (I) the wrong variety of
poison having been gathered, (2) breach of a taboo, (3) witch- ')
1!
craft, (4) anger of the owners of the forest where the creeper II
grows, (5) age ofthe poison, (6) anger of the ghosts, (7) sorcery, 'ji
(8) use. i;
I,
'j
If at its first seance the oracle kills fowls without discrimina- ~l;
tion, slaying one after the other without sparing a single one,
:;;1
they say that it is 'foolish' poison. More often it happens at "I
li
'I.
seances thatthe poison fails to affect the fowls and they say that I,'
it is 'weak poison' or 'dead poison'. If some four medium-sized ,I'
';:~i
,h
fowls are in succession unaffected by the poison they stop the
,:'
ft!
,~
seance, and later the poison will be thrown away; sin'ce once
it has lost its potency there are no means of restoring it, whereas ~
ill
ifit is over-potent it may, after being kept for some time, become II
good, and by this Azande mean discriminating. Sometimes
when the fowls appear totally unaffected by the poison they
~
administer the usual doses to one ofthem while asking the oracle
the straightforward question, 'Ifyou are good oracle poison kill
this fowl. If you are worthless oracle poison spare it.' If the
poison is 'good poison' or 'strong poison' it can demonstrate
its potency forthwith.
The poison may be over-potent because the gatherers col-
lected it from the wrong kind of creeper, for there are two
varieties of poison creeper, that called nawada and that called
andegi. The andegi kills fowls without regard to the questions put
to it. It-is unnecessary to seek a cause, for people know at once
by its action that it is andegi and they wrap it up in leaves and
place it in hiding and wait some months for it to 'cool'. If at
the end of this time it is still 'stupid' they either throw it away
15 6 Oracles
or seek to discover whether witchcraft or some other cause is
now responsible for its failure to give correct judgements.
The explanation of why poison kills all the fowls by reference
to andegi is only adduced when the poison is freshly gathered
and being tested to determine its worth. If a packet of poison
is passed as good nawada and at a later seance kills all the fowls
some other explanation must be sought, and its behaviour is
usually attributed to witchcraft.
If at its preliminary test or at any later test the poison is
impotent and does not kill a single fowl Azande generally attri-
bute its behaviour to breach of a taboo. Today when poison
is often purchased from Azande of the Congo there is grave
danger of it having been polluted by someone through whose
hands it has passed, and once it has come into contact with
an unclean person it is permanently ruined.
Witchcraft is often cited as a cause for wrong verdicts. It also
may render the oracle impotent, though impotency is usually
attributed to breach oftaboo. Generally speaking, the presence
of witchcraft is shown by the oracle killing two fowls in answer
to the same question, or in sparing two fowls in answer to the
same question when it has killed a fowl at the same seance. In
such cases the poison is evidently potent and its failure to give
correct judgements may be due to a passing influence of witch-
craft. For the time being the seance may be stopped and
resumed on another day when it is hoped that witchcraft will
no longer be operative. Nevertheless, unless the oracle makes
many consecutive errors Azande do not generally close the'
seance, because it often happens that witchcraft interferes with
the working of the poison in relation to a single and particular
question, and in no way influences it in relation to other ques-
tions. The witch is preventing the oracle from giving an accu-
rate reply to a certain question that concerns him but is not
seeking to interfere in questions that do not concern him nor
to destroy the poison completely.
Sometimes the poison refuses to function properly on a cer-
tain day because the operator is in an unlucky state, 'his condi-
tion is bad', as Azande say, and this means that there is witch-
craft about him and by coming into contact with the oracle
poison he has transmitted the ill-luck to it, so that the 'condi-
tion' of the oracle is bad likewise. Sometimes they interrupt
Oracles 157
their questions to ask the oracle whether it is being troubled
by witchcraft, and people say that it may then kill a fowl, after
having shown itself unable to do so before, or spare a fowl, after
having killed all the previous ones, in order to inform the ques-
tioner that there is witchcraft present. A man does not ask one
packet of poison whether another packet is good.
Ifat its first testing after it has been gathered the oracle poison l
fails to operate, and the man who gathered it is certain that
he kept the taboos required of him and that it did not come
,
into contact with any polluting influence, its impotency may I"
be attributed to the anger of the owners of the soil where it
was dug up. Or it may be said that some foreigner must have
polluted the poison, unknown to the gatherer, while the party ;.;
,[
were on their return journey. Such explanations are, however,
seldom offered and would seldom be accepted. The man who I
puts them forward wishes to excuse himselffrom responsibility. i
One sometimes hears it said that a packet of poison has lost
its power because it has been kept too long. Men have, however,
denied to me that this is possible, asserting that breach of taboo,
or witchcraft, or some other cause must be responsible for loss
of strength.
It is said that occasionally the ghosts are held responsible.
Men say that if a man gathers oracle poison in the Congo and
neglects to give part of it to his father as first-fruits the ghosts 11
may corrupt it. .
H
I'
Finally, any poison will lose its power with use. A man gener-
ally prepares for a seance more poison than will be used in the l~ I
,J)
tests. At the end of the seance he gathers up what is left and "II
'~
stores it apart from unused poison. Poison can be used at least
twice and, ifit is of good quality, sometimes three or four times. ~
I',"
Sometimes they prepare a mixture of fresh and used poison. ,:l1
At length its strength is exhausted. Azande know this happens 'ij%j
'1'
and they merely say 'It is exhausted' without advancing any
mystical cause for its loss of potency.
Sometimes the poison acts in a peculiar manner inside the
fowl and experience is necessary to interpret correctly its re-
actions; It sometimes happens that a fowl appears to have sur-
vived its ordeal but dies later when it is running about in the
grass, or even after its owner has brought it back to his home-
stead. I have never observed a fowl revive after it has appeared
II
\
158 Oracles
to fall lifeless to the ground, but I was told that this occasionally
occurs. Indeed, I have heard Mbira boast of having addressed
an apparently lifeless hen for a long time with such vehemence
and good sense that it finally survived. When such things
happen young Azande do not always know how to interpret
them, but old and experienced men are seldom at a loss to
explain the fowl's behaviour. People do not care to act on a
verdict of the oracle unless it is given without ambiguity.
If a fowl collapses very slowly and then suddenly recovers
this means that there is some evil influence hanging over the
operator. 'His condition is bad.' fowls may die slowly in a long
series of spasms as though the poison were uncertain whether
to kill them or not, and this probably means that witchcraft
is trying to influence the oracle.
The oracle must reply to the question in either an affirmative
or a negative, but sometimes it sees more than it is asked and
wants to let the people know what it has seen, e.g. they may
ask it whether a man will be bewitched ifhe goes ajourney, and
the oracle knows that although he will not be bewitched his
family will be bewitched during his absence or that he himself
will be attacked by sorcery. Or they may ask whether a certain
man will fall ill this month, and the oracle sees that although
he will be in good health this month he will fall sick next month.
It tries to tell people these facts and at the same time to answer
the questions put to it.
v
It will have been noted that Azande act experimentally within
the framework of their mystical notions. They act as we would
have to act if we had no means of making chemical and physio-
logical analyses and we wanted to obtain the same results as
they want to obtain. As soon as the poison is brought back from
its forest home it is tested to discover whether some fowls will
live and others die under its influence. It would be unreasonable
to use poison without first having ascertained that all fowls to
which it is administered do not die or do not live. The oracle
would then be a farce. Each seance must be in itself experiment-
ally consistent. Thus if the first three fowls survive Azande will
always be apprehensive. They at once suspect that the oracle
is not working properly. But ifthen,afterwards, the fourth fowl
Oracles 159
dies, they are content. They will say to you, 'You see the poison
is good, it has spared the first three fowls but it has killed this
one.' Zande behaviour, though ritual, is consistent, and the
reasons they give for their behaviour, though mystical, are in-
tellectually coherent.
If their mystical notions allowed them to generalize their
observations they would perceive, as we do, that their faith is l!
without foundations. They themselves provide all the proof
necessary. They say that they sometimes test new poison or old
poison which they fear has been corrupted by asking it silly
questions. At full moon they administer the poison to a fowl
and address it thus:
,
Poison oracle, tell the chicken about those two spears over there.
As I am about to go up to the sky, if I will spear the moon today
with my spears, kill the fowl. If! will not spear the moon today, poison
oracle spare the fowl.
If the oracle kills the fowl they know that it is corrupt.
And yet Azande do not see that their oracles tell them noth-
'iI
ing! Their blindness is not due to stupidity: they reason excel-
lently in the idiom of their beliefs, but they cannot reason out-
side, or against, their beliefs because they have no other idiom
in which to express their thoughts.
The reader will naturally wonder what Azande say when :,1'
"J
subsequent events prove the prophecies of the poison oracle to I,:
be wrong. Here again Azande are not surprised at such an out- '"
come, but it does not prove to them that the oracle is futile.
nrl
It rather proves how well founded are their beliefs in witchcraft ,~I
and sorcery and taboos. The contradiction between what the
oracle said would happen and what actually has happened is :~
just as glaring to Zande eyes as it is to ours, but they never ~
for.a moment question the virtue of the oracle in general but iW
,!~~
seek only to account for the inaccuracy of this particular poison.
Moreover, even if the oracle was not deflected from the
straight path of prophecy by witchcraft or bad magic there are
other reasons which would equally account for its failure. It
may be that the particular venture about the success of which
a man was consulting the oracle was not at the time of consulta-
tion threatened by witchcraft, but that a witch intervened at
some time between the consultation and the commencement >~i
of the undertaking. ,",1
't
"~'11:
~
11"
':
160 Oracles
Azande see as well as we that the failure of their oracle to
prophesy truly calls for explanation, but so entangled are they
in mystical notions that they must make use ofthem to account
for the failure. The contradiction between experience and one
mystical notion is explained by reference to other mystical
notions.
Normally there is little chance of the oracle being proved
wrong, for it is usually asked questions to which its answers can-
not well be challenged by subsequent experience, since the in-
quirer accepts the verdict and does not seek to check it by ex-
periment. Thus were a man to ask the oracle, 'If I build my
homestead in such-and-such a place will I die there?' or, 'If
my son is sponsored by so-and-so in the circumcision ceremonies
will he die?' and were the oracle to reply 'Yes' to either of these
queries, he would not construct his homestead in the ill-omened
place nor allow his son to be sponsored by the inauspicious man.
Consequently he would never know what would have hap-
pened it he had not taken the advice of the oracle. Also, the
verdict of the oracle is usually in accordance with the workings
of nature, and were a man to receive the reply that it is safe
for him to marry a certain girl because she will not die within
the next few years, or that he is assured of his harvest of eleusine
if he sows it in a certain spot in the bush, there would be little
likelihood of the oracle being proved wrong, as the chances of
the girl dying or of the hardy eleusine being totally destroyed
would be small.
Furthermore, only certain types of question are regularly put'
to the oracle: questions relating to witchcraft, sickness, death,
lengthyjourneys, mourning and vengeance, changing of home-
stead sites, lengthy agricultural and hunting enterprises, and
so forth. One does not ask the poison oracle about small matters
or questions involving minute precision with regard to time.
A man would not ask such a question as: 'Will I kill a bush-
buck if I go hunting tomorrow?' and since men do not ask
that sort ofquestion they do not receive immediat~ detailed in-
structions which might go amiss and expose the falsity of the
oracle.
Indeed, as a rule Aiande do not ask questions to which
answers are easily tested by experience and they ask only
those questions which embrace contingencies. The answers
Oracles 161
either cannot be tested, or if proved by subsequent events to
be erroneous permit an explanation of the error. In the last
resort errors can always be explained by attributing them to
mystical interference. But there is no need to suppose that the
Zande is conscious of an evasion of clear issues. In restricting
his questions to certain well-known types he is conforming to
tradition. It does not occur to him to test the oracle experiment-
ally unless he has grave suspicions about a particular packet
of poison.
Moreover, the mai~E~ose of theor~cle liesjIl its ability
to reveal th~phY of I!lysti~<l.L[orc~s. When Az~_nd~l>0ut
health or marri'!g~--9r hunting they are seeking information
aboiitth~mQY~J!le!lt of psychic forces which might cause them
mis~tune. They do not attempt simply to discover tlieo6]eC=
tive conditions at a certain point of time in the future, nor the
objective results ofa certain action, but the inclination of mysti-
cal powers, for these conditions and result depend upon di.em.
Azande envisage a future;ari-lriaividual's future that is to-say,
depe-nd~ni:-iip.Qn_rnystical forces. Hence when the oracle paints
a black horizon for a man he is glad to have been warned
because now that he knows the dispositions of witchcraft he can
get into touch with it and have the future ch<inged_ !()~be mwe
favourable to him.
By means of his oracles a Zande can discover the mystical
forces which hang over a man and doom him in advance, and ..;
having diSCOverea tIiem necan counteract them or alter his
plans to avoid the doom which awaits him in any particular
Ii
venture. Hence it is evident that the answers he receives do not '.j
generally concern objective happenings-aria- therefore cannot K,
eil
easily be contra-ry to experIence. \1;,
None the less, I have often noticed that Azande on being in- ,.,
formed that sickness lies ahead of them do not even proceed u
to discover the name of the witch whose influence is going to
cause them sickness and get him to blowout water but merely
wait for a few days and then consult the oracle again to find
out whether their health will be good for the coming month,
hoping that by the time ofthe second consultation the evil influ- t~1
ence which hung over their future at the time of the first con-
sultation will no longer be there.
It follows that present and future have not entirely the same
r62 Oracles
meaning for Azande as they have for us. It is difficult to formu-
late the problem in our language, but it would appear from
their behaviour that the present and future overlap in some way
so that the present partakes of the future as it were. Hence a
man's future health and happiness depend on future conditions
that are already in existence and can be exposed by the oracles
and altered. The future depends on the disposition of mystical
forces that can be tackled here and now. Moreoever, when the
oracles announce that a man will fall sick, i.e. be bewitched
in the near future, his 'condition' is therefore already bad, his
future is already part of him. Azande cannot explain these mat-
ters, they content themselves with believing and enacting them.
By the same token, the oracle is protected by its position in
the order of events. When a Zande wishes to slay a witch who
has killed one of his kinsmen or a thief who has stolen his prop-
erty he does not ask the oracle to identify the witch or thief
and then make magic against this known person, but he first
makes magic against an unknown criminal, and when people
in the neighbourhood die he asks the oracle whether one of
them is the victim of his punitive magic.
But in spite of the many ways in which belief in the poison
oracle is sustained it may be doubted whether it could have
maintained prestige in a democratic community. In Zandeland
its verdicts derive an historic sanction from the fact that its ver-
dicts were traditionally backed by the full authority of the king.
The decisions of the king's oracle were final. Had there been'
any appeal from this to private oracles there would have been'
general confusion, since everybody would have been able to
produce oracular verdicts to support his own point of view and
there would have been no way of deciding between them. In
legal disputes, therefore, the authority ofthe poison oracle was
formerly the authority of the king, and this in itselfwould tend
to prevent any serious challenge to its veracity.
vr
There is a final problem to discuss. As I have recorded in earlier
sections, each situation demands the particular pattern of
thought appropriate to it. Hence an individual in one situation
will employ a notion he excludes in a different situation. The
many beliefs I have recorded are so many different tools of
Oracles r63
thought, and he selects the ones that are chiefly to his advantage.
A Zande does not readily accept an oracular verdict which con-
flicts seriously with his interests. No one believes that the oracle
is nonsense, but everyone thinks that for some particular reason
in this particular case the particular poison used is in error in
respect to himself. Azande are only sceptical of particular I:I
oracles and not of oracles in general, and their scepticism is
~"
always expressed in a mystical idiom that vouches for the
validity of the poison oracle as an institution.
Also, apart from criminal cases, there can be no doubt that
a man takes advantage of every loop-hole the oracle allows him
to obtain what he wants or to refrain from doing what he does
not want to do. Moreover, he uses the authority of the oracle
to excuse his conduct or to compel others to accept it. The oracle
is often very useful in such a question as whether a man's wife
shall pay her parents a visit. It is difficult for the husband to
forbid her visit, but ifhe can say that the oracles advise against
ifhe can both prevent it and checkmate objections on the part
of his parents-in-law.
In the actual·operation of the oracle, Azande like to receive
a favourable prediction in the first test and to put off the corro~
borative test that may contradict it for as long as possible.
Tradition allows them a certain latitude in the order in' which
they arrange their. questions to the oracle and also in the
number of doses that are administered to the fowls. There is
an art in questioning the oracle, for it must answer 'yes' or 'no'
to a question and a man can therefore define the terms of the 'I'
answer by stating them in the question. By close interpretations
of the reactions of fowls to the poison it is often possible to
qualify the declaration oracles give by killing or sparing them.
In all this Azande are not employing trickery. A man uses
for his individual needs in certain situations those notions that
most favour his desires. Azande cannot go beyond the limits
'"I
...
;
"
:J
set by their culture and invent notions, but within these limits '~
human behaviour is not rigidly determined by custom and a
man has some freedom of action and thought.
:~,
~'l
:1

CHAPTER X
Other Zande Oracles
I
AZANDE esteem dakpa, or the termites oracle, next to the poison
oracle. A man will not place a verdict of the termites oracle
before the rubbing-board oracle for confirmation, and he will
not place a verdict of the poison oracle to the termites for con-
firmation. If more than one oracle is consulted they consult
always the lesser before the greater in the order of: (I) rubbing-
board, (2) termites, (3) poison. Dakpa is the poor man's poison
oracle. There are no expenses involved, for a man has only to
find a termite mound and insert two branches of different trees
into one of their runs and return next day to see which of the
two the termites have eaten. The main drawback to the oracle
from the Zande point of view is that it is lengthy and limited.
It takes an entire night to answer a question, and very few ques-
tions can be asked at the same time.
On all important matters the decisions of the termites oracle
must be corroborated by the poison oracle. Legal action cannot
be taken without a decision from the poison oracle. But poison
is expensive and it is cheaper to obtain preliminary verdicts
from the termites and ask the poison oracle for a final decision.
Thus a man finds out which among half a dozen sites is a suit-'
able one for him to build his homestead in and can place the
choice ofthe termites before the poison oracle for confirmation.
Women can consult the termites oracle as well as men, and
children sometimes use it. It is known to, and can be used by,
everyone.
The oracle is regarded as very reliable, much more so than
the rubbing-board. Azande say that the termites do not listen
to all the talk which is going on outside in homesteads and only
hear the questions put to them. Older men try to consult the
termites oracle at the beginning of each month to discover
whether they will continue in good health. A wealthy man asks
the same question of the poison oracle.
Oracles 16 5
The oracle is called after one of the trees, a branch of which
is inserted into a termite mound in its operation. Azande may
address their questions to these branches. Nevertheless, they
ordinarily address the termites and in their commentaries on
the oracle it is clear that they think of the termites as listening
to their questions and giving answers to them. But the fact that
they address both shows that no general and independent intel- i:
ligence is attributed to either the termites or to the trees but·
only a specific intelligence in the operation of the oracle, and
that it is the oracle as a whole, as something sui generis, which
is the object of inquiry.
A man ought to observe the same taboos as for the poison
oracle, but they are less strict. The termites are always
approached towards evening. A man goes to one of his own
termite mounds because people may object ifhe disturbs their
termites by thrusting branches into their runs. He does not take
branches of dakpa and kpoyo from his homestead because these
two trees are found everywhere in the bush. With the haft of
his spear he opens up one of the great shafts that lead into the
mound, or one of the runs at the side of it, takes a branch of
either tree in each hand, and, speaking to the termites, which
rush to the seat of disturbance, says .SOri'fe such words as: '0
termites, I will die this year, eat dakpa. I will not die, eat kpoyo.'
He may address the branches as though they were eating:
'Dakpa I will die this year, dakpa you eat; I will not die this
year, kpoyo you eat.' The words vary according to the question,
but they are always spoken in one of the traditional forms.
While making this speech he thrusts the two branches into the
shaft, or run, and after placing a few of the lumps of earth he
has excavated around them he returns home.
Early next morning the questioner goes to the mound to
receive an answer. The termites may have eaten kpoyo and left ~:) ,j
dakpa, or they may have eaten dakpa and left kpoyo. They may :'J
have eaten both, or left both untouched. The answer depends ',I
therefore on the way in which the question is phrased. When :1
.!'
it concerns the welfare of the question,er or of his kin it is asked
in such a way that if the termites eat dakpa it is a prophecy
of misfortune, and if they eat kpoyo it is a prophecy of good for- i"llW!
';},
tune. Having obtained a verdict, they can then either place it
at once before the poison oracle or, if they do not want to do
,~"
166 Oracles
that, they can obtain from termites a gingo, or corroborative
test, similar to that they demand from the poison oracle.' I
believe, however, that this is not generally done, for cases which
demand such care are usually so important that they must go
before the poison oracle which will supply all the confirmation
needed.
It sometimes happens that neither branch is eaten. Then
Azande simply say that the termites refuse an answer and they
try another mound. Frequently they eat both branches. This
is not an invalid verdict as it is when the poison kills or spares
two fowls. The same unambiguous answer is not expected from
termites. Doubtless we would be right in relating this accept-
ance of unprecise answers to the fact that the questions put to
the termites oracle have not the social importance of those put
to the poison oracle and do not settle legal issues. When both
branches are eaten the interpretation is not a complete answer
to the question but a partial answer, e.g if dakpa is the branch
mainly eaten, the answer is a qualified verdict leaning towards
a negative or ,,!:ffirmative according to the terms of the question.
If both branches are eaten about equally Azande may say
that the ants were merely hungry and ate to satisfy their appe-
tites, or they may say that a taboo has been broken or that
witchcraft has interfered with the oracle. But they do not evoke
mystical entities to account for the failure of the termites oracle
to give unambiguous answers with the frequency and lux-
uriance of those evoked to explain discrepancies in the verdicts
of the poison oracle.
II
Another Zande oracle is called mapingo. It can be used byevery-
one, but except in choosing a site for a homestead adult men
do not often employ it. It is considered especially the oracle
of women and children. It is by using mapingo that children
gain their first experience in oracular consultations. Its opera-
tion is as simple as can be. Three small pieces of wood about
half an inch long and rounded are cut from branches of a tree
or from the stalk of the manioc. Three pieces of wood are re-
quired for each question to be asked. The material always lies
ready at hand. Two pieces of stick are placed side by side on the
ground and the third piece is placed on top of them and parallel
Oracles 16 7
to them. These little pieces of stick are generally arranged just
before nightfall in a clearing at the edge of the garden where
it borders the homestead or at the back of a hut. When asking
about a new homestead they are arranged in a small clearing
in the proposed site in the bush. The oracle gives its answers
by the three sticks remaining in position all night or by the
structure falling. Azande sometimes say that a person ought to
!:
observe the usual taboos for a short time before using the three
sticks oracle, but I very much doubt whether anyone does so.
When the sticks have been set in position a man addresses
them and tells them what he wants enlightenment about, or
perhaps, we might rather say, he speaks a conditional clause
over them.
When consulting about a homestead site a man usually erects
two piles of sticks, one for himself and one for his wife. He
addresses his pile somewhat as follows:
I will die, there is badness over this homestead site, if T build my ':i
'I
homestead on it I will die there. Mapingo you scatter to show that
my 'condition' is bad. I will not die on it, let me come to examine
I
you and find you in position to show my good 'condition'.
Generally the question is so phrased that the displacement
I
of the sticks gives an inauspicious prognostication and their
remaining in position gives an auspicious prognostication.
The oracle is not considered important. Women and children
ask it many questions, but they are questions about their own
affairs and have little social significance. Men also use it on
occasion. Its verdicts are not made public and a man cannot
approach a witch on its findings alone. It is sometimes used
as a preliminary to the termites or poison oracles. N otwith-
standing, it is considered very reliable, especially in reference
to homestead sites, and a man would not neglect its advice. ,
'~,
III
The most used of all Zande oracles is iwa, the rubbing-board.
.,!)
:<i
The poison oracle needs preparation. Often, especially today, II
y
it is difficult to obtain oracle poison, and a man may have to
wait many days until he learns that a kinsman or blood-brother .1 ,1
1
is about to consult the oracle and will allow him to bring one
or two fowls to solve his problems. But a man cannot wait when
he fears that he may be a victim of witchcraft or trickery. At
,~"
168 Oracles
any time a sudden problem may confront him, a sudden suspi-
cion assail him. If he possesses a rubbing-board oracle and is
qualified to use it he will carry it with him wherever he goes
in his little skin or grass plaited bag so that he can take it out
at a moment's notice and inquire from it what he is to do.
Otherwise he may easily find a kinsman or friend who will con-
sult their oracles on his behalf, for it is a small service and costs
them nothing. And it is not only in situations requiring imme-
diate action that the rubbing-board is more suitable an oracle
than the poison, but also in dozens of situations when the issues
are of minor importance and hardly worthy of being presented
to the poison oracle. Azande do not place complete faith in
its statements and contrast its reliability unfavourably with that
of the other oracles which have been described. They put its
revelations on a par with those of witch-doctors. The rubbing-
board is looked upon as an inferior judge which sorts out a case
so thatit is reduced to preliminary issues that can then go before
the poison oracle. Thus a man is ill and a great many persons
occur to him as likely to be bewitching him. It would be a
tedious and expensive business to place six or seven names
before the poison oracle when, perhaps, the last on the
list is the right one. But it will not take him longer than ten
minutes to place the names before his little wooden instrument,
and when it has chosen from among them the responsible witch
all that need be asked of the greater poison oracle is to confirm
its choice. The poison oracle is always the final authority, and
if the matter is one involving relations between two persons if
must be consulted. For this reason, unless the matter is urgent,
they bring all important social questions directly before the
poison oracle. It is only minor or preliminary questions that
are asked ofthe rubbirtg-board. Azande say it answers so many
questions that it is bound to be wrong sometimes. We may
observe that this admission can be made because situations of
use are minor and do not involve social interrelations.
The oracles consist of miniature table-like constructions.
Smaller ones are carried about in bags. Larger ones are kept
in homesteads. They are carved out ofthe wood of various trees.
They have two parts, the 'female', or the flat surface of the table
supported by two legs and its tail, and the 'male', or the piece
which fits the surface of the table like a lid. The shape of the
Oracles 169
table is round to oval. When not in operation a barkcloth cover-
ing is tied over the head of the instrument.
When fashioning a rubbing-board a man is subject to taboos.
He must abstain for two days from sexual relations and from
the same foods prohibited in connexion with the poison oracle
before he commences to manufacture it. He cuts it with an adze,.
fashioning the bottom part before the upper part. He then
blackens it by rubbing the surface with a red-hot spear. The
carving of the board is only part of the process of manufacture.
It is still nothing but two pieces of carved wood and has to be
endowed with mystical potency, i.e. the wood has to be trans-
formed into an oracle. This is done by two actions. In the first
place the table is anointed with medicine derived from roots j
which have been boiled, their juices then mixed with oil and
boiled again and, during this second boiling, stirred and
addressed in the pot. I was t()ld that the owner says over the
pot:
I
J
This is my rubbing-board oracle which I am going to doctor. When
I consult it on a man's behalf may it speak the truth, may it foretell
the death (threatened death) of a man. May it reveal things to me,
may it not hide things from me. May it not lose its potency. If a man
eats tabooed food, such as elephant (and comes near my oracle), may
it not lose its potency.
He then takes the mixture off the fire and, having rnade in-
cisions on the table of the oracle, he rubs some of it into them.
The remainder of the oil andjuices he mixes with ashes ofvari-
ous plants and rubs them on to the face of the table. The in-
cisions may be partly the cause of the lid of the oracle sticking
or running smoothly on the table according to the direction
of pressure.
In the second place the oracle has to be buried. It has been
doctored, but the medicines have to be given time to sink in
and there is still 'coldness' about it which must be removed.
It is wrapped up in new barkcloth or perhaps in the skin of
a small animal like a small bushbuck and is placed in a hole
dug in the centre of a path. The earth is well trodden down
to disguise the fact that something has been buried there,
because if a man notices that the earth has been disturbed he :"'\"
will go round the spot in fear of sorcery, and this will spoil the
preparation of the oracle, because it is passers-by who 'take
170 Oracles
away all "coldness" from the rubbing-board in the centre of
the path' as they pass over it. After two days the owner digs
it up.
He now tests it by rubbing the wooden lid backwards and
forwards on the table. He says to it: 'Rubbing-board, if you
will speak the truth to people, stick.' It sticks in declaration
of its potency and powers of discrimination. The owner then
addresses the oracle, saying, 'Rubbing-board, I take a little
wealth to redeem you with it. You speak the truth to me. I
take ashes to hold your legs with them. You speak the truth
to me.' He then places a knife before it as a payment. Since
the knife is taken away again Azande say, 'He deceives the rub-
bing~board with a knife.' He then binds barkcloth round it and
places it under his veranda. The rubbing-board is ready for use.
It is operated in the following manner. A man sits on the
ground and steadies the board by placing his right foot on its
tail, while with his right hand he jerks the lid backwards and
forwards, towards and away from him, between his thumb and
first finger. Before operating the oracle he squeezes juices of
plants or grates wood ofvarious trees on to the table. Generally
they use the fruit ofthe Kaffir apple for this purpose. The opera-
tor dips the lid into a gourd of water which he keeps at his side
and applies its flat surface to the surface of the table. As soon
as they touch, the juices or gratings on the table become mois-
tened and begin to froth and bubble. He jerks the lid backwards
and forwards a few times and then begins to question the oracle.
From time to time during the consultations he moistens the lid'
in the gourd of water.
When the operator jerks the lid over the table it generally
either moves smoothly backwards and forwards or it sticks to
the board so firmly that no jerking will further move it, and
it has to be pulled upwards with considerable force to detach
it from the table. These two actions-smooth sliding and firm
sticking-are the two ways in which the oracle answers ques-
tions. They correspond to the slaying or sparing offowls by the
poison, the eating or refusing of the branches by the termites,
and the disturbance or. non-disturbance of the pile of sticks.
Every question is therefore framed thus: If such is the case,
'stick', and if such is not the case, 'run smoothly'. In consulta-
tions ofthe rubbing-board sticking ofthe lid almost always gives
Oracles 17 1
an affirmative answer and smooth running of the lid almost
always gives a negative answer.
Whatever other questions a man intends to place before the
oracle he gerierally asks as his first question, 'Shall I die this
year?' and the oracle runs smoothly, giving its answer 'No'.
Sometimes instead of going smoothly backwards and forwards
or sticking fast the lid runs from side to side or round and round.
Sometimes it alternately sticks and runs. The oracle is here
refusing to give a verdict, and this generally means that it is
doubtful of the issue or sees something outside the terms of the
question that would seriously qualify the unequivocal answer
given by either sticking or sliding.
Strictly speaking, as with the poison and termites oracles, a
second and confirmatory test should be made. If the lid has
stuck in the first test, thenin the second test it must slide back-
wards and forwards smoothly, and vice versa, if the verdict is
to be valid. In fact, however, they very seldom make a second
test. In important issues the question will be placed before the
poison oracle, which supplies all the confirmation needed. Also
Azande must be aware that the second test always confirms the
first one. But they do not trouble themselves to any great extent
about such matters because in serious questions a higher auth-
ority is consulted.
IV
Before consulting the rubbing-board its owner is supposed
to observe the same prohibitions as those in force when using
the poison oracle, though it is not required that he shall observe
them for so long a period before operating it. Since the oracle
may be consulted at a moment's notice the taboos would prove
irksome to its owner and to those who wish him to operate it
on their behalf if it were not that their observance can be
rendered unnecessary by a process known as 'spoiling the rub-
bing-board'. A piece of an elephant's skin, or a fish-bone, and
perhaps of a piece of wood on which a woman has sat (for a
menstruating woman can destroy the potency of any oracle if
she goes near it), are burnt and the ashes are rubbed over the
table of the board. Instead of burning fish-bone they may
sprinkle the table with a few drops of water in which a fish has
been cooked. It will not matter after this has been done if a
"~/
 17 2 Oracles
man eats elephant's flesh or fish or a menstruating woman
approaches the oracle.
That people, in fact, do not observe taboos is well known.
Azande have told me that whilst every oracle owner sleeps regu-
larly with his wife, few have been heard to refuse to operate
the oracle on that account. Nevertheless, they say that a sincere
man who wished to keep his oracle potent would not use it for
two or three days after having had sexual intercourse. They
attribute much of the error in the oracle's judgements to slack-
ness in this respect. In the past only a few old men owned rub-
bing-boards, and in those days taboos were more rigidly
observed, for old men are more careful than their juniors to
avoid contamination. Even today not many men own rubbing-
boards.
Azande say that the accuracy of a rubbing-board depends
upon its not becoming 'cold'. They say that if a man's oracle
makes many mistakes he will realize that it has lost its potency.
I t can be rehabilitated by placing medicine on its table and
wrapping it in barkcloth and burying it again in a path. I was
told that they say to it as they place it in the hole, 'You are
rubbing-board, why do you lie? Speak the truth.' Mter two
days the owner digs it up and burns a little benge and rubs the
soot on to the board, and says to it, 'Rubbing-board you speak
the truthjust as benge speaks it.' He then puts a pinch of oracle
poison on the table, wraps it up in backcloth, and places it
under his veranda to rest for a few days.
Operation of the rubbing-board differs from operation of the '
other oracles in that only certain persons can operate it. With
the exception of the one or two peculiar women who have even
been known to consult the poison oracle and a few female ghost-
diviners, these persons are all middle-aged or old men. Women
may occasionally watch the oracle being operated as it is being
consulted in homesteads, and often publicly, unlike the poison
and termites oracles which are consulted in the bush, but they
are not encouraged to approach near to it-and cannot operate it.
Children do not use it, and I have never known a young man
operate it. Moreover, its use is not conditioned by age and sex
alone, for it is necessary for a man to have absorbed certain
medicines before he can hope to operate the instrument. The
poison, termites, and three sticks oracles can be operated by
I
~
Oracles 173
any man who has kept the taboos, but when the rubbing-board
is operated the owner as well as the instrument itself has to be
doctored. He will get a magician to doctor him. The potency
of the oracle is due to the medicines which it absorbs when the
board is being made, the medicines applied to its table before
use, the medicines ru bbed into the hand and foot of the operator
and eaten by him, and its operation subject to customary condi-
tions.
Only the owner of a rubbing-board uses it. He will not let
other people operate it. He will consult it about the affairs of
'kinsmen and great friends without exacting a fee, but from
neighbours he expects a present of a knife, or half-piastre, or
ring, or some such small gift. He can courteously demand
payment by pointing out that the oracle will not work properly
unless it sees a gift laid on the ground before it. If you do not
produce a fee when asking him to consult his oracle on your
behalf he may say that he is sorry but his rubbing-board is
broken or that he has not kept taboos the day before or that
he has not ritually cleansed himself after assisting at a burial.
Azande are well aware that people can cheat in operating
the rubbing-board oracle, and this is one of the reasons why.
they consider it inferior to the other oracles. However, they do
not think that people often cheat, and a man only mentions
that an operator may have cheated when the oracle has spoken
against him or he particularly dislikes the operator. No owner
of a good oracle cheats or fails to observe taboos lest it cease
to be a good oracle. Some men's oracles have a wide reputation
for accuracy and enjoy this reputation in contrast to others.
Since the rubbing-board has no legal status, there is no reason
why tradition and authority should exclude, or explain away
by assertion and by the use ofsecondary elaborations, the possi-
bility ofimproper manipulation. A man must believe, or at any
rate express belief, in the poison oracle and submit to its
declarations. But the statements of the rubbing-board need not
inconvenience anyone except its consulter, and custom does
not compel a man to use it or to submit to its verdicts.
I have little doubt that the operator improperly manipulates
the oracle in most inquiries. Nevertheless, owners of rubbing-
boards frequently consult them about their own affairs, and it
can scarcely be imagined that they deliberately cheat on such
174 Oracles Oracles 175
occasions. It may also be asked why, if they cheat, they should them to dream true dreams. When the dream is oracular it will
go to the trouble of burying the board and doctoring it and then prophesy the future truly and warn a man of impending
themselves. danger and tell him offortune to come, e.g. ifhe goes hunting
It must be difficult for a man who is considering a question or pays a visit to his prince to ask for a gift of spears he will
to move his hand quite haphazardly when the movement is sup- kill animals or receive the gift. If the dream is a nightmare he
posed to provide an answer to the question and when pressure will then be abletosee the features of the witch who is attacking
makes all the difference between 'Yes' and 'No'. It may well him so that he will know his enemy.
be that Azande are not entirely aware that they control the Azande attach great importance to the prognostication of
oracle in accordance with conclusions reached in their minds dreams. They say that dream-prophecies are as true as those
and that between the thinking out of the questions and the ofthe rubbing-board oracle. However, dreams do not often lead
movement of the hand in answer the middle clause, 'I must directly to action. Azande like to place their prognostications
make the lid stick (or go smoothly)" -is not consciously formu- before one of the four main oracles, benge, dakpa, mapingo, and
lated. If this is the case 'cheating' is perhaps too strong a word iwa, to make certain that they have correctly interpreted them.l
to use.
1 For further information on dreams, cf. Appendix II.
v
Azande speak ofdreams as oracles, for they reveal hidden things
(soroka). In a sense all dreams foretell events, but some more
clearly than others. Those dreams in which a man actually ex-
periences witchcraft portend misfortune to the dreamer as a
consequence of his having been bewitched, and dreams about
ghosts, not recorded in this book, inform people about happen-
ings among the dead. But many dreams are explained solely
in terms of prophecy without reference to witchcraft, though
what may at the time have appeared to h'!ve been a dream
of one type may be shown by events to have been a dream of
another type and to have been misinterpreted. '
There are stereotyped explanations of dreams. These are
generally straightforward affirmations that what happened in
the dream will later take place in waking life, but sometimes
dream images are regarded as symbols which require inter-
pretation. Nevertheless, in such cases the interpretation is often
traditional and it is merely necessary to find someone who
knows it. Azande do not always know how to interpret dreams,
though an obscure dream is always vaguely considered good
or bad. On the whole, whatwe would call bad dreams are evi-
dence of witchcraft, and what we would call pleasant dreams
are oracular and the dreamer believes that they may happen
to him in the future.
Some men eat ngua musumo, dream-medicines, which enable